Mercurial > repos > jeremyjliu > region_motif_enrichment
changeset 17:7afdfd4f4c1b draft
Uploaded
line wrap: on
line diff
Binary file region-motif-compare/._Galaxy-Workflow-H3k27ac_Peaks_Minus_H3k4me3_Peaks_From_bigBeds.ga has changed
Binary file region-motif-compare/._Galaxy-Workflow-Region_Motif_Count_Comparison_Combined_Motifs.ga has changed
Binary file region-motif-compare/._Galaxy-Workflow-Region_Motif_Count_Comparison_Jaspar_Jolma_Motifs.ga has changed
Binary file region-motif-compare/._Galaxy-Workflow-Region_Motif_Count_Comparison_Mouse_Motifs.ga has changed
Binary file region-motif-compare/._Galaxy-Workflow-Region_Motif_Count_Comparison_Pouya_Motifs.ga has changed
Binary file region-motif-compare/._Galaxy-Workflow-Region_Motif_Count_Comparison_Test_Motifs.ga has changed
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/Galaxy-Workflow-H3k27ac_Peaks_Minus_H3k4me3_Peaks_From_bigBeds.ga Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,145 @@ +{ + "a_galaxy_workflow": "true", + "annotation": "{\n \"refinery_type\": \"analysis\",\n \"refinery_relationships\": [\n {\n \"category\": \"1-1\",\n \"set1\": \"k27ac bigBed\",\n \"set2\": \"k4me3 bigBed\"\n }\n ]\n}", + "format-version": "0.1", + "name": "H3k27ac Peaks Minus H3k4me3 Peaks From bigBeds", + "steps": { + "0": { + "annotation": "", + "id": 0, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "k27ac bigBed" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 145, + "top": 238 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"k27ac bigBed\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "1": { + "annotation": "", + "id": 1, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "k4me3 bigBed" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 142, + "top": 396 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"k4me3 bigBed\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "2": { + "annotation": "", + "id": 2, + "input_connections": { + "input": { + "id": 0, + "output_name": "output" + } + }, + "inputs": [], + "name": "SAM/BAM/GFF/BedGraph/BigBed/VCF to Bed", + "outputs": [ + { + "name": "output", + "type": "bed" + } + ], + "position": { + "left": 358, + "top": 177 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "toolshed.g2.bx.psu.edu/repos/timpalpant/java_genomics_toolkit/IntervalToBed/1.0.0", + "tool_state": "{\"input\": \"null\", \"__rerun_remap_job_id__\": null, \"chromInfo\": \"\\\"/Users/jeremyliu1/galaxy-dist/tool-data/shared/ucsc/chrom/hg19.len\\\"\", \"__page__\": 0}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "3": { + "annotation": "", + "id": 3, + "input_connections": { + "input": { + "id": 1, + "output_name": "output" + } + }, + "inputs": [], + "name": "SAM/BAM/GFF/BedGraph/BigBed/VCF to Bed", + "outputs": [ + { + "name": "output", + "type": "bed" + } + ], + "position": { + "left": 363, + "top": 379 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "toolshed.g2.bx.psu.edu/repos/timpalpant/java_genomics_toolkit/IntervalToBed/1.0.0", + "tool_state": "{\"input\": \"null\", \"__rerun_remap_job_id__\": null, \"chromInfo\": \"\\\"/Users/jeremyliu1/galaxy-dist/tool-data/shared/ucsc/chrom/hg19.len\\\"\", \"__page__\": 0}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "4": { + "annotation": "{\n \"output\": {\n \"name\": \"K27ac Filtered Enhancers\",\n \"description\": \"Histone k27ac modifications enhanced region sites, with k4me3 region site intersections removed.\",\n \"type\": \"bed\"\n }\n}", + "id": 4, + "input_connections": { + "input1": { + "id": 2, + "output_name": "output" + }, + "input2": { + "id": 3, + "output_name": "output" + } + }, + "inputs": [], + "name": "Subtract", + "outputs": [ + { + "name": "output", + "type": "input" + } + ], + "position": { + "left": 827, + "top": 298 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "toolshed.g2.bx.psu.edu/repos/devteam/subtract/gops_subtract_1/1.0.0", + "tool_state": "{\"input2\": \"null\", \"__page__\": 0, \"input1\": \"null\", \"min\": \"\\\"1\\\"\", \"__rerun_remap_job_id__\": null, \"returntype\": \"\\\"\\\"\", \"chromInfo\": \"\\\"/Users/jeremyliu1/galaxy-dist/tool-data/shared/ucsc/chrom/hg19.len\\\"\"}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + } + } +} \ No newline at end of file
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/Galaxy-Workflow-Region_Motif_Count_Comparison.ga Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,149 @@ +{ + "a_galaxy_workflow": "true", + "annotation": "{\n \"refinery_type\": \"analysis\",\n \"refinery_relationships\": [\n {\n \"category\": \"1-1\",\n \"set1\": \"Region 1\",\n \"set2\": \"Region 2\"\n }\n ]\n}", + "format-version": "0.1", + "name": "Region Motif Count Comparison", + "steps": { + "0": { + "annotation": "", + "id": 0, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 1" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 207, + "top": 226 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 1\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "1": { + "annotation": "", + "id": 1, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 2" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 206, + "top": 360 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 2\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "2": { + "annotation": "{\n \"out_tab\": {\n \"name\": \"motifs_count_region_1\",\n \"description\": \"Motif counts in region 1.\",\n \"type\": \"tabular\"\n }\n}", + "id": 2, + "input_connections": { + "in_bed": { + "id": 0, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 404, + "top": 207 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"t\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "3": { + "annotation": "{\n \"out_tab\": {\n \"name\": \"motifs_count_region_2\",\n \"description\": \"Motif counts in region 2.\",\n \"type\": \"tabular\"\n }\n}", + "id": 3, + "input_connections": { + "in_bed": { + "id": 1, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 420, + "top": 354 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"t\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "4": { + "annotation": "{\n \"out_enriched\": {\n \"name\": \"motifs_enriched\",\n \"description\": \"Motifs enriched in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n },\n \"out_depleted\": {\n \"name\": \"motifs_depleted\",\n \"description\": \"Motifs depleted in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n }\n}", + "id": 4, + "input_connections": { + "in_tab_1": { + "id": 2, + "output_name": "out_tab" + }, + "in_tab_2": { + "id": 3, + "output_name": "out_tab" + } + }, + "inputs": [], + "name": "Region Motif Count Compare", + "outputs": [ + { + "name": "out_enriched", + "type": "tabular" + }, + { + "name": "out_depleted", + "type": "tabular" + } + ], + "position": { + "left": 637.5, + "top": 252 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_compare", + "tool_state": "{\"in_tab_1\": \"null\", \"db_type\": \"\\\"t\\\"\", \"__rerun_remap_job_id__\": null, \"in_tab_2\": \"null\", \"__page__\": 0}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + } + } +} \ No newline at end of file
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/Galaxy-Workflow-Region_Motif_Count_Comparison_Combined_Motifs.ga Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,277 @@ +{ + "a_galaxy_workflow": "true", + "annotation": "{\n \"refinery_type\": \"analysis\",\n \"refinery_relationships\": [\n {\n \"category\": \"1-1\",\n \"set1\": \"Region 1\",\n \"set2\": \"Region 2\"\n }\n ]\n}", + "format-version": "0.1", + "name": "Region Motif Count Comparison Combined Motifs", + "steps": { + "0": { + "annotation": "", + "id": 0, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 1" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 153.73957872390747, + "top": 193.73959398269653 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 1\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "1": { + "annotation": "", + "id": 1, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 2" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 158.68751573562622, + "top": 417.69794511795044 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 2\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "2": { + "annotation": "", + "id": 2, + "input_connections": { + "in_bed": { + "id": 0, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 391.7256932258606, + "top": 144.71181535720825 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"p\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "3": { + "annotation": "", + "id": 3, + "input_connections": { + "in_bed": { + "id": 0, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 385.6146092414856, + "top": 253.63542985916138 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"j\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "4": { + "annotation": "", + "id": 4, + "input_connections": { + "in_bed": { + "id": 1, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 390.7187352180481, + "top": 369.72919511795044 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"p\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "5": { + "annotation": "", + "id": 5, + "input_connections": { + "in_bed": { + "id": 1, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 393.6874852180481, + "top": 490.7014241218567 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"j\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "6": { + "annotation": "", + "id": 6, + "input_connections": { + "input1": { + "id": 2, + "output_name": "out_tab" + }, + "queries_0|input2": { + "id": 3, + "output_name": "out_tab" + } + }, + "inputs": [], + "name": "Concatenate datasets", + "outputs": [ + { + "name": "out_file1", + "type": "input" + } + ], + "position": { + "left": 649.4340672492981, + "top": 198.9652943611145 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "cat1", + "tool_state": "{\"__page__\": 0, \"__rerun_remap_job_id__\": null, \"input1\": \"null\", \"queries\": \"[{\\\"input2\\\": null, \\\"__index__\\\": 0}]\"}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "7": { + "annotation": "", + "id": 7, + "input_connections": { + "input1": { + "id": 4, + "output_name": "out_tab" + }, + "queries_0|input2": { + "id": 5, + "output_name": "out_tab" + } + }, + "inputs": [], + "name": "Concatenate datasets", + "outputs": [ + { + "name": "out_file1", + "type": "input" + } + ], + "position": { + "left": 654.4687962532043, + "top": 410.99655961990356 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "cat1", + "tool_state": "{\"__page__\": 0, \"__rerun_remap_job_id__\": null, \"input1\": \"null\", \"queries\": \"[{\\\"input2\\\": null, \\\"__index__\\\": 0}]\"}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "8": { + "annotation": "{\n \"out_enriched\": {\n \"name\": \"motifs_enriched\",\n \"description\": \"Motifs enriched in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n },\n \"out_depleted\": {\n \"name\": \"motifs_depleted\",\n \"description\": \"Motifs depleted in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n },\n \"out_plots\": {\n \"name\": \"motif_counts_plots\",\n \"description\": \"Summary plots for motifs in regions 1 versus regions 2.\",\n \"type\": \"png\"\n }\n}", + "id": 8, + "input_connections": { + "in_tab_1": { + "id": 6, + "output_name": "out_file1" + }, + "in_tab_2": { + "id": 7, + "output_name": "out_file1" + } + }, + "inputs": [], + "name": "Region Motif Count Compare", + "outputs": [ + { + "name": "out_enriched", + "type": "tabular" + }, + { + "name": "out_depleted", + "type": "tabular" + }, + { + "name": "out_plots", + "type": "png" + } + ], + "position": { + "left": 904.2639007568359, + "top": 285.75696563720703 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_compare", + "tool_state": "{\"in_tab_1\": \"null\", \"db_type\": \"\\\"c\\\"\", \"__rerun_remap_job_id__\": null, \"in_tab_2\": \"null\", \"__page__\": 0}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + } + } +} \ No newline at end of file
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/Galaxy-Workflow-Region_Motif_Count_Comparison_Jaspar_Jolma_Motifs.ga Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,153 @@ +{ + "a_galaxy_workflow": "true", + "annotation": "{ \"refinery_type\": \"analysis\", \"refinery_relationships\": [ { \"category\": \"1-1\", \"set1\": \"Region 1\", \"set2\": \"Region 2\" } ] }", + "format-version": "0.1", + "name": "Region Motif Count Comparison Jaspar Jolma Motifs", + "steps": { + "0": { + "annotation": "", + "id": 0, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 1" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 199.86806535720825, + "top": 215.927086353302 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 1\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "1": { + "annotation": "", + "id": 1, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 2" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 197.85417985916138, + "top": 392.95835161209106 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 2\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "2": { + "annotation": "{\n \"out_tab\": {\n \"name\": \"motifs_count_region_1\",\n \"description\": \"Motif counts in region 1.\",\n \"type\": \"tabular\"\n }\n}", + "id": 2, + "input_connections": { + "in_bed": { + "id": 0, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 402.1771168708801, + "top": 199.79862260818481 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"j\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "3": { + "annotation": "{\n \"out_tab\": {\n \"name\": \"motifs_count_region_2\",\n \"description\": \"Motif counts in region 2.\",\n \"type\": \"tabular\"\n }\n}", + "id": 3, + "input_connections": { + "in_bed": { + "id": 1, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 413.1666798591614, + "top": 390.7882161140442 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"j\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "4": { + "annotation": "{\n \"out_enriched\": {\n \"name\": \"motifs_enriched\",\n \"description\": \"Motifs enriched in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n },\n \"out_depleted\": {\n \"name\": \"motifs_depleted\",\n \"description\": \"Motifs depleted in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n },\n \"out_plots\": {\n \"name\": \"motif_counts_plots\",\n \"description\": \"Summary plots for motifs in regions 1 versus regions 2.\",\n \"type\": \"png\"\n }\n}", + "id": 4, + "input_connections": { + "in_tab_1": { + "id": 2, + "output_name": "out_tab" + }, + "in_tab_2": { + "id": 3, + "output_name": "out_tab" + } + }, + "inputs": [], + "name": "Region Motif Count Compare", + "outputs": [ + { + "name": "out_enriched", + "type": "tabular" + }, + { + "name": "out_depleted", + "type": "tabular" + }, + { + "name": "out_plots", + "type": "png" + } + ], + "position": { + "left": 643.6042098999023, + "top": 283.6909866333008 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_compare", + "tool_state": "{\"in_tab_1\": \"null\", \"db_type\": \"\\\"j\\\"\", \"__rerun_remap_job_id__\": null, \"in_tab_2\": \"null\", \"__page__\": 0}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + } + } +} \ No newline at end of file
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/Galaxy-Workflow-Region_Motif_Count_Comparison_Mouse_Motifs.ga Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,153 @@ +{ + "a_galaxy_workflow": "true", + "annotation": "{\n \"refinery_type\": \"analysis\",\n \"refinery_relationships\": [\n {\n \"category\": \"1-1\",\n \"set1\": \"Region 1\",\n \"set2\": \"Region 2\"\n }\n ]\n}", + "format-version": "0.1", + "name": "Region Motif Count Comparison Mouse Motifs", + "steps": { + "0": { + "annotation": "", + "id": 0, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 1" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 201, + "top": 135 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 1\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "1": { + "annotation": "", + "id": 1, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 2" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 196, + "top": 331 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 2\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "2": { + "annotation": "{\n \"out_tab\": {\n \"name\": \"motifs_count_region_1\",\n \"description\": \"Motif counts in region 1.\",\n \"type\": \"tabular\"\n }\n}", + "id": 2, + "input_connections": { + "in_bed": { + "id": 0, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 411, + "top": 122 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"m\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "3": { + "annotation": "{\n \"out_tab\": {\n \"name\": \"motifs_count_region_2\",\n \"description\": \"Motif counts in region 2.\",\n \"type\": \"tabular\"\n }\n}", + "id": 3, + "input_connections": { + "in_bed": { + "id": 1, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 408, + "top": 322 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"m\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "4": { + "annotation": "{\n \"out_enriched\": {\n \"name\": \"motifs_enriched\",\n \"description\": \"Motifs enriched in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n },\n \"out_depleted\": {\n \"name\": \"motifs_depleted\",\n \"description\": \"Motifs depleted in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n },\n \"out_plots\": {\n \"name\": \"motif_counts_plots\",\n \"description\": \"Summary plots for motifs in regions 1 versus regions 2.\",\n \"type\": \"png\"\n }\n}", + "id": 4, + "input_connections": { + "in_tab_1": { + "id": 2, + "output_name": "out_tab" + }, + "in_tab_2": { + "id": 3, + "output_name": "out_tab" + } + }, + "inputs": [], + "name": "Region Motif Count Compare", + "outputs": [ + { + "name": "out_enriched", + "type": "tabular" + }, + { + "name": "out_depleted", + "type": "tabular" + }, + { + "name": "out_plots", + "type": "png" + } + ], + "position": { + "left": 692.5, + "top": 206 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_compare", + "tool_state": "{\"in_tab_1\": \"null\", \"db_type\": \"\\\"m\\\"\", \"__rerun_remap_job_id__\": null, \"in_tab_2\": \"null\", \"__page__\": 0}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + } + } +} \ No newline at end of file
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/Galaxy-Workflow-Region_Motif_Count_Comparison_Pouya_Motifs.ga Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,153 @@ +{ + "a_galaxy_workflow": "true", + "annotation": "{\n \"refinery_type\": \"analysis\",\n \"refinery_relationships\": [\n {\n \"category\": \"1-1\",\n \"set1\": \"Region 1\",\n \"set2\": \"Region 2\"\n }\n ]\n}", + "format-version": "0.1", + "name": "Region Motif Count Comparison Pouya Motifs", + "steps": { + "0": { + "annotation": "", + "id": 0, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 1" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 206.98611497879028, + "top": 225.97918057441711 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 1\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "1": { + "annotation": "", + "id": 1, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 2" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 205.9791874885559, + "top": 359.98958706855774 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 2\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "2": { + "annotation": "{\n \"out_tab\": {\n \"name\": \"motifs_count_region_1\",\n \"description\": \"Motif counts in region 1.\",\n \"type\": \"tabular\"\n }\n}", + "id": 2, + "input_connections": { + "in_bed": { + "id": 0, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 403.98266649246216, + "top": 206.98612332344055 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"p\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "3": { + "annotation": "{\n \"out_tab\": {\n \"name\": \"motifs_count_region_2\",\n \"description\": \"Motif counts in region 2.\",\n \"type\": \"tabular\"\n }\n}", + "id": 3, + "input_connections": { + "in_bed": { + "id": 1, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 423.98266649246216, + "top": 351.9861385822296 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"p\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "4": { + "annotation": "{\n \"out_enriched\": {\n \"name\": \"motifs_enriched\",\n \"description\": \"Motifs enriched in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n },\n \"out_depleted\": {\n \"name\": \"motifs_depleted\",\n \"description\": \"Motifs depleted in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n },\n \"out_plots\": {\n \"name\": \"motif_counts_plots\",\n \"description\": \"Summary plots for motifs in regions 1 versus regions 2.\",\n \"type\": \"png\"\n }\n}", + "id": 4, + "input_connections": { + "in_tab_1": { + "id": 2, + "output_name": "out_tab" + }, + "in_tab_2": { + "id": 3, + "output_name": "out_tab" + } + }, + "inputs": [], + "name": "Region Motif Count Compare", + "outputs": [ + { + "name": "out_enriched", + "type": "tabular" + }, + { + "name": "out_depleted", + "type": "tabular" + }, + { + "name": "out_plots", + "type": "png" + } + ], + "position": { + "left": 637.4931030273438, + "top": 251.98958659172058 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_compare", + "tool_state": "{\"in_tab_1\": \"null\", \"db_type\": \"\\\"p\\\"\", \"__rerun_remap_job_id__\": null, \"in_tab_2\": \"null\", \"__page__\": 0}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + } + } +} \ No newline at end of file
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/Galaxy-Workflow-Region_Motif_Count_Comparison_Test_Motifs.ga Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,153 @@ +{ + "a_galaxy_workflow": "true", + "annotation": "{\n \"refinery_type\": \"analysis\",\n \"refinery_relationships\": [\n {\n \"category\": \"1-1\",\n \"set1\": \"Region 1\",\n \"set2\": \"Region 2\"\n }\n ]\n}", + "format-version": "0.1", + "name": "Region Motif Count Comparison Combined Motifs", + "steps": { + "0": { + "annotation": "", + "id": 0, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 1" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 206.95139360427856, + "top": 225.94445848464966 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 1\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "1": { + "annotation": "", + "id": 1, + "input_connections": {}, + "inputs": [ + { + "description": "", + "name": "Region 2" + } + ], + "name": "Input dataset", + "outputs": [], + "position": { + "left": 205.94445085525513, + "top": 359.95488023757935 + }, + "tool_errors": null, + "tool_id": null, + "tool_state": "{\"name\": \"Region 2\"}", + "tool_version": null, + "type": "data_input", + "user_outputs": [] + }, + "2": { + "annotation": "{\n \"out_tab\": {\n \"name\": \"motifs_count_region_1\",\n \"description\": \"Motif counts in region 1.\",\n \"type\": \"tabular\"\n }\n}", + "id": 2, + "input_connections": { + "in_bed": { + "id": 0, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 403.94792222976685, + "top": 206.9514012336731 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"t\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "3": { + "annotation": "{\n \"out_tab\": {\n \"name\": \"motifs_count_region_2\",\n \"description\": \"Motif counts in region 2.\",\n \"type\": \"tabular\"\n }\n}", + "id": 3, + "input_connections": { + "in_bed": { + "id": 1, + "output_name": "output" + } + }, + "inputs": [], + "name": "Region Motif Intersect", + "outputs": [ + { + "name": "out_tab", + "type": "tabular" + } + ], + "position": { + "left": 423.94792222976685, + "top": 351.9514012336731 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_intersect", + "tool_state": "{\"__page__\": 0, \"db_type\": \"\\\"t\\\"\", \"in_bed\": \"null\", \"__rerun_remap_job_id__\": null}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + }, + "4": { + "annotation": "{\n \"out_enriched\": {\n \"name\": \"motifs_enriched\",\n \"description\": \"Motifs enriched in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n },\n \"out_depleted\": {\n \"name\": \"motifs_depleted\",\n \"description\": \"Motifs depleted in regions 1 versus regions 2.\",\n \"type\": \"tabular\"\n },\n \"out_plots\": {\n \"name\": \"motif_counts_plots\",\n \"description\": \"Summary plots for motifs in regions 1 versus regions 2.\",\n \"type\": \"png\"\n }\n}", + "id": 4, + "input_connections": { + "in_tab_1": { + "id": 2, + "output_name": "out_tab" + }, + "in_tab_2": { + "id": 3, + "output_name": "out_tab" + } + }, + "inputs": [], + "name": "Region Motif Count Compare", + "outputs": [ + { + "name": "out_enriched", + "type": "tabular" + }, + { + "name": "out_depleted", + "type": "tabular" + }, + { + "name": "out_plots", + "type": "png" + } + ], + "position": { + "left": 644.4549102783203, + "top": 257.9444580078125 + }, + "post_job_actions": {}, + "tool_errors": null, + "tool_id": "region_motif_compare", + "tool_state": "{\"in_tab_1\": \"null\", \"db_type\": \"\\\"t\\\"\", \"__rerun_remap_job_id__\": null, \"in_tab_2\": \"null\", \"__page__\": 0}", + "tool_version": "1.0.0", + "type": "tool", + "user_outputs": [] + } + } +} \ No newline at end of file
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/README.md Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,204 @@ +# Region-Motif-Compare Tools +Version 1.1 Released 2014 +Park Laboratory +Center for Biomedical Informatics +Harvard University + +Contact +Jeremy Liu (jeremy.liu@yale.edu) +Nils Gehlenborg (nils@hms.harvard.edu) + +## Overview +### Structure +The tool suite consists of: + +1. Two Rscripts: region_motif_compare.r and region_motif_intersect.r +2. Two Xml Files: region_motif_compare.xml and region_motif_intersect.xml +3. Motif Database Directory: region_motif_db +4. Dependency Library Directory: region_motif_lib +5. Galaxy Workflows: Files with suffix ".ga" that can be imported into the local +Galaxy instance after installation of the tool. + +### Description +1. **region_motif_intersect.r** (1 bed -> 1 tsv): +Takes one bed file of regions as input. Then it calculates +the number of intersections of the regions and the motifs. region_motifs_intersect.r +outputs a tab separated values (tsv) file of motif names and intersection counts. +**Important Note:** region_motif_intersect.r makes no assumptions about the nature +of the input regions. For example, if overlapping regions are inputted, motifs that +intersect the overlap will be double counted. Thus, it is recommended that regions +be merged before using this tool, using the merge tool in the Galaxy toolshed. + +2. **region_motif_compare.r** (2 tsv -> 2 tsv & 1 png): +Takes as input two tsv files of motifs / regions intersection +counts. These generally originate from running region_motif_intersect.r on two sets +of different regions with the same query motif database. Based on the counts, +region_motif_compare.r then determines the enrichment (or depletion) of certain +motifs across the two regions. This is done by a correcting for the size and gc +content of the region, and applying a Poisson test to the counts. +Then, region_motif_compare.r outputs the most significant enriched or depleted +motifs as a tsv. In addition, the tool outputs a diagnostic plot containing +graphical representations of the motif counts, gc correction curves, and significant +motifs that distinguish the two regions (selected via p value). + +3. **region_motif_db**: Contains motif positions as compressed, indexed tabix files. + +4. **region_motif_lib**: Contains dependencies (i.e. plotting.r) for region_motif_compare.r + +## Installation +Directions for installing the region-motif-compare tools into a personal computer +and a local Galaxy instance. + +1. Follow the online directions to install a local instance of Galaxy (getgalaxy.org). +Optionally, follow the directions to install Refinery (refinery-platform.readthedocs.org) + +2. Clone the github repository to your local computer + ```` + git clone https://github.com/parklab/refinery-galaxy-tools.git + cd refinery-galaxy-tools/region-motif-compare + ```` + +3. Make a directory for the tools in Galaxy instance. This serves as a category +for the tool in the tools sidebar. You can also place the tools in an existing +or alternatively named directory, but remember to update tool_conf.xml to reflect this. + ```` + cd ~/galaxy-dist/tools/ + mkdir my_tools + cd my_tools + ```` + +4. Copy over ".r" and ".xml" files, as well as `region_motif_db` and `region_motif_lib` + ```` + cd refinery-galaxy-tools/region-motif-compare + cp *.r ~/galaxy-dist/tools/my_tools + cp *.xml ~/galaxy-dist/tools/my_tools + cp -r region_motif_db ~/galaxy-dist/tools/my_tools + cp -r region_motif_lib ~/galaxy-dist/tools/my_tools + ```` + +5. Edit `~/galaxy-dist/tool_conf.xml` to reflect the addition of the new tools. +Add the following lines within the `<toolbox>` tags. If in Step 3 you copied +the tools to a different directory than `my_tools`, edit the code snippet +to reflect the correct path name. + ```` + <section id="mTools" name="My Tools"> + <tool file="my_tools/region_motif_intersect.xml" /> + <tool file="my_tools/region_motif_compare.xml" /> + </section> + ```` + +6. Download the motif databases and place them into `region_motif_db` + ```` + cd ~/galaxy-dist/tools/my_tools/region_motif_db + wget ????/pouya_motifs.bed.bgz + wget ????/pouya_motifs.bed.bgz.tbi + wget ????/jaspar_jolma_motifs.bed.bgz + wget ????/jaspar_jolma_motifs.bed.bgz.tbi + wget ????/mm9_motifs.bed.bgz + wget ????/mm9_motifs.bed.bgz.tbi + ```` + +7. Install the Bioconductor R package Rsamtools for dealing with tabix files + ``` + $ R + > source("http://bioconductor.org/biocLite.R") + > biocLite("Rsamtools") + ```` + +8. If in Step 3 you copied the tools to an existing directory or an alternatively +named directory, you must edit the following file paths. + In `region_motif_intersect.r` and `region_motif_compare.r` edit `commonDir`: + ```` + # Replace this line + commonDir = concat(workingDir, "/tools/my_tools") + # With this edited line + commonDir = concat(workingDir, "<relative_path_from_galaxy_root>/<tool_directory>") + ```` + In addition, edit `region_motif_intersect.xml` and `region_motif_compare.xml` to + reflect the path of the tools relative to the galaxy root directory. + ```` + <command interpreter="bash"> + /usr/bin/R --slave --vanilla -f $GALAXY_ROOT_DIR/<path_to_tools>/region_motif_intersect.r --args $GALAXY_ROOT_DIR $db_type $in_bed $out_tab + </command> + ```` + ```` + <command interpreter="bash"> + /usr/bin/R --slave --vanilla -f $GALAXY_ROOT_DIR/<path_to_tools>/region_motif_compare.r --args $GALAXY_ROOT_DIR $db_type $in_tab_1 $in_tab_2 $out_enriched $out_depleted $out_plots + </command> + ```` + +## Running the Tools +### Running from Galaxy +1. To run the tools as workflows, import the .ga workflows included in the github +via the Galaxy workflow user interface. Then, upload and select two input BED files. + +2. To run the tools individually, select the tool from the tools toolbar, provide +a BED file (Region Motif Intersect) or two tsv files (Region Motif Compare), and +select a query database from the dropdown menu. + +### Running from Refinery +1. Import the .ga workflows into a local Galaxy instance. These workflows have +already been annotated for Refinery. + +2. Add the local Galaxy instance to the Refinery installation. + ```` + python manage.py create_workflowengine <instance_id> "<group_name>" + ```` + +3. Import the Galaxy workflows into Refinery. + ```` + python manage.py import_workflows + ```` +4. Run the tools from the Refinery user interface. + +### Running as Command Line Tools +You can also run the tools from the command line, an example of which is shown below. +More information is found in the headers of the r source files. +```` +cd ~/galaxy-dist/tools/my_tools +R --slave --vanilla -f region_motif_intersect.r --args ~/galaxy-dist p <path_to_bed_file> <path_to_output_tsv> +R --slave --vanilla -f region_motif_compare.r --args ~/galaxy-dist p <path_to_region1_counts> <path_to_region2_counts> <enriched_motifs_output_tsv> <depleted_motifs_output_tsv> <plots_png> +```` + +## Interpreting Results +### Motif Database and Result Notation +TF motif positions for hg19 and mm9 were curated from three databases: +ENCODE TF motif database "Pouya" (http://compbio.mit.edu/encode-motifs/) +JASPAR database "Jaspar" (http://jaspar.genereg.net/) +DNA binding specificities of human transciption factors "Jolma" (http://www.ncbi.nlm.nih.gov/pubmed/23332764) + +For ENCODE TF motifs, the genomic locations were taken straight from the database. +In addition, position weight matrices (pwms) were obtained by averaging the +sites in the genome for a motif. These are labeled with "\_8mer\_". +Fake motifs were also generated, by shuffling the pwms of actual motifs and +mapping to the genome and are labeled with "_8mer_C". + +For JASPAR and Jolma motifs, mast was run to determine genomic locations from the +provided pwms. The motif alignmment thresholds were set to the top 5k, 20k, 100k, and +250k sites and the redundant maps removed with the top 30k sites have the same score. +These are labeled with "_t5000" and likewise. + + +## Motif Tabix File Creation +Starting with a BED file of motif positions (minimal chr, start, end), follow +below to generate a tabix file that can be placed in `region_motif_db` and +used by the tools. + +1. Download Tabix (http://sourceforge.net/projects/samtools/files/tabix/) and install. +Add `tabix` and `bgzip` binaries to your file path. + ```` +tar -xvjf tabix-0.2.6.tar.bz2 +cd tabix-0.2.6 +make + ```` + +2. Construct bgzip files and index files. + ```` +cd ~/galaxy-dist/tools/my_tools/region_motif/db +(grep ^"#" jaspar_motifs.bed; grep -v ^"#" jaspar_motifs.bed | sort -k1,1 -k2,2n) | bgzip > jaspa_motifs.bed.bgz +tabix -p bed jaspar_motifs.bed.bgz # this generates jaspar_motifs.bed.bgz.tbi + ```` + +3. Add the path to `jaspar_motifs.bed.bgz` to the selection options for the variable +`motifDB` in `region_motif_intersect.r` and `region_motif_compare.r`. To enable +the new database in Galaxy, you will have to edit the xml files for both tools.
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/region_motif_compare.r Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,188 @@ +# Name: region_motif_compare.r +# Description: Reads in two count files and determines enriched and depleted +# motifs (or any location based feature) based on poisson tests and gc +# corrections. All enrichment ratios relative to overall count / gc ratios. +# Author: Jeremy liu +# Email: jeremy.liu@yale.edu +# Date: 14/07/03 +# Note: This script is meant to be invoked with the following command +# R --slave --vanilla -f ./region_motif_compare.r --args <workingdir> <db> <intab1> <intab2> +# <enriched_tab> <depleted_tab> <plots_png> +# <workingdir> is working directory of galaxy installation +# <db> types: "t" test, "p" pouya, "j" jaspar jolma, "m" mouse, "c" combined +# Dependencies: plotting.r + +# Auxiliary function to concatenate multiple strings +concat <- function(...) { + input_list <- list(...) + return(paste(input_list, sep="", collapse="")) +} + +# Supress all warning messages to prevent Galaxy treating warnings as errors +options(warn=-1) + +# Set common and data directories +args <- commandArgs() +workingDir = args[7] +dbDir = concat(workingDir, "/region_motif_db") +dbCode = args[8] +# dbCode "c" implemented when pwmFile is loaded +if (dbCode == "t" | dbCode == "p") { + pwmFile = concat(dbDir, "/pwms/pouya.pwms.from.seq.RData") +} else if (dbCode == "j") { + pwmFile = concat(dbDir, "/pwms/jaspar.jolma.pwms.from.seq.RData") +} else if (dbCode == "m") { + pwmFile = concat(dbDir, "/pwms/mm9.pwms.from.seq.RData") +} else if (dbCode == "c") { # rest of dbCode "c" implemeted when pwmFile loaded + pwmFile = concat(dbDir, "/pwms/pouya.pwms.from.seq.RData") + pwmFile2 = concat(dbDir, "/pwms/jaspar.jolma.pwms.from.seq.RData") +} else { + pwmFile = concat(dbDir, "/pwms/pouya.pwms.from.seq.RData") +} + +# Set input and reference files +inTab1 = args[9] +inTab2 = args[10] +enrichTab = args[11] +depleteTab = args[12] +plotsPng = args[13] + +# Load dependencies +source(concat(workingDir, "/region_motif_lib/plotting.r")) + +# Auxiliary function to read in tab file and prepare the data +read_tsv <- function(file) { + data = read.table(file, sep="\t", stringsAsFactors=FALSE) + names(data)[names(data) == "V1"] = "motif" + names(data)[names(data) == "V2"] = "counts" + return(data) +} + +startTime = Sys.time() +cat("Running ... Started at:", format(startTime, "%a %b %d %X %Y"), "...\n") + +# Loading motif position weight matrix (pwm) file and input tab file +cat("Loading and reading input region motif count files...\n") +load(pwmFile) # pwms data structure +if (dbCode == "c") { # Remaining implementation of dbCode "c" combined + temp = pwms + load(pwmFile2) + pwms = append(temp, pwms) +} +region1DF = read_tsv(inTab1) +region2DF = read_tsv(inTab2) +region1Counts = region1DF$counts +region2Counts = region2DF$counts +names(region1Counts) = region1DF$motif +names(region2Counts) = region2DF$motif + +# Processing count vectors to account for missing 0 count motifs, then sorting +cat("Performing 0 count correction and sorting...\n") +allNames = union(names(region1Counts), names(region2Counts)) +region1Diff = setdiff(allNames, names(region1Counts)) +region2Diff = setdiff(allNames, names(region2Counts)) +addCounts1 = rep(0, length(region1Diff)) +addCounts2 = rep(0, length(region2Diff)) +names(addCounts1) = region1Diff +names(addCounts2) = region2Diff +newCounts1 = append(region1Counts, addCounts1) +newCounts2 = append(region2Counts, addCounts2) +region1Counts = newCounts1[sort.int(names(newCounts1), index.return=TRUE)$ix] +region2Counts = newCounts2[sort.int(names(newCounts2), index.return=TRUE)$ix] + +# Generate gc content matrix +gc = sapply(pwms, function(i) mean(i[2:3,3:18])) + +# Apply poisson test, calculate p and q values, and filter significant results +cat("Applying poisson test...\n") +rValue = sum(region2Counts) / sum(region1Counts) +pValue = sapply(seq(along=region1Counts), function(i) { + poisson.test(c(region1Counts[i], region2Counts[i]), r=1/rValue)$p.value +}) +qValue = p.adjust(pValue, "fdr") +indices = which(qValue<0.1 & abs(log2(region1Counts/region2Counts/rValue))>log2(1.5)) + +# Setting up output diagnostic plots, 4 in 1 png image +png(plotsPng, width=800, height=800) +xlab = "region1_count" +ylab = "region2_count" +lim = c(0.5, 5000) +layout(matrix(1:4, ncol=2)) +par(mar=c(5, 5, 5, 1)) + +# Plot all motif counts along the linear correlation coefficient +plot.scatter(region1Counts+0.5, region2Counts+0.5, log="xy", xlab=xlab, ylab=ylab, + cex.lab=2.2, cex.axis=1.8, xlim=lim, ylim=lim*rValue) +abline(0, rValue, untf=T) +abline(0, rValue*2, untf=T, lty=2) +abline(0, rValue/2, untf=T, lty=2) + +# Plot enriched and depleted motifs in red, housed in second plot +plot.scatter(region1Counts+0.5, region2Counts+0.5, log="xy", xlab=xlab, ylab=ylab, + cex.lab=2.2, cex.axis=1.8, xlim=lim, ylim=lim*rValue) +points(region1Counts[indices]+0.5, region2Counts[indices]+0.5, col="red") +abline(0, rValue, untf=T) +abline(0, rValue*2, untf=T, lty=2) +abline(0, rValue/2, untf=T, lty=2) + +# Apply and plot gc correction and loess curve +cat("Applying gc correction, rerunning poisson test...\n") +ind = which(region1Counts>5) +gc = gc[names(region2Counts)] # Reorder the indices of pwms to match input data +lo = plot.scatter(gc,log2(region2Counts/region1Counts),draw.loess=T, + xlab="gc content of motif",ylab=paste("log2(",ylab,"/",xlab,")"), + cex.lab=2.2,cex.axis=1.8,ind=ind) # This function is in plotting.r +gcCorrection = 2^approx(lo$loess,xout=gc,rule=2)$y + +# Recalculate p and q values, and filter for significant entries +pValueGC = sapply(seq(along=region1Counts),function(i) { + poisson.test(c(region1Counts[i],region2Counts[i]),r=1/gcCorrection[i])$p.value +}) +qValueGC=p.adjust(pValueGC,"fdr") +indicesGC = which(qValueGC<0.1 & abs(log2(region1Counts/region2Counts*gcCorrection))>log2(1.5)) + +# Plot gc corrected motif counts +plot.scatter(region1Counts+0.5, (region2Counts+0.5)/gcCorrection, log="xy", + xlab=xlab, ylab=paste(ylab,"(normalized)"), cex.lab=2.2, cex.axis=1.8, + xlim=lim, ylim=lim) +points(region1Counts[indicesGC]+0.5, + (region2Counts[indicesGC]+0.5)/gcCorrection[indicesGC], col="red") +abline(0,1) +abline(0,1*2,untf=T,lty=2) +abline(0,1/2,untf=T,lty=2) + +# Trim results, compile statistics and output to file +# Only does so if significant results are computed +if(length(indicesGC) > 0) { + # Calculate expected counts and enrichment ratios + cat("Calculating statistics...\n") + nullExpect = region1Counts * gcCorrection + enrichment = region2Counts / nullExpect + + # Reorder selected indices in ascending pvalue + cat("Reordering by ascending pvalue...\n") + indicesReorder = indicesGC[order(pValueGC[indicesGC])] + + # Combine data into one data frame and output to two files + cat("Splitting and outputting data...\n") + outDF = data.frame(motif=names(pValueGC), p=as.numeric(pValueGC), q=qValueGC, + stringsAsFactors=F, region_1_count=region1Counts, + null_expectation=round(nullExpect,2), region_2_count=region2Counts, + enrichment=enrichment)[indicesReorder,] + names(outDF)[which(names(outDF)=="region_1_count")]=xlab + names(outDF)[which(names(outDF)=="region_2_count")]=ylab + indicesEnrich = which(outDF$enrichment>1) + indicesDeplete = which(outDF$enrichment<1) + outDF$enrichment = ifelse(outDF$enrichment>1, + round(outDF$enrichment,3), + paste("1/",round(1/outDF$enrichment,3))) + write.table(outDF[indicesEnrich,], file=enrichTab, quote=FALSE, + sep="\t", append=FALSE, row.names=FALSE, col.names=TRUE) + write.table(outDF[indicesDeplete,], file=depleteTab, quote=FALSE, + sep="\t", append=FALSE, row.names=FALSE, col.names=TRUE) +} + +# Catch display messages and output timing information +catchMessage = dev.off() +cat("Done. Job started at:", format(startTime, "%a %b %d %X %Y."), + "Job ended at:", format(Sys.time(), "%a %b %d %X %Y."), "\n")
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/region_motif_compare.xml Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,32 @@ +<tool id="region_motif_compare" name="Region Motif Count Compare"> + <description>for comparing the motif counts in different region sets</description> + <requirements> + <requirement type="set_environment">RMOTIF_PATH</requirement> + </requirements> + <command interpreter="Rscript"> + region_motif_compare.r --args \$RMOTIF_PATH $db_type $in_tab_1 $in_tab_2 $out_enriched $out_depleted $out_plots + </command> + <inputs> + <param name="in_tab_1" type="data" format="tabular" label="Region Set 1 Motif Count File"/> + <param name="in_tab_2" type="data" format="tabular" label="Region Set 2 Motif Count File"/> + <param name="db_type" type="select" label="Select Motif Database" > + <option value="t">Test Pouya Subset (hg19)</option> + <option value="p">Pouya Encode Motifs (hg19)</option> + <option value="j">Jaspar and Jolma Motifs (hg19)</option> + <option value="m">Mouse Motifs (mm9)</option> + <option value="c">Pouya, Jaspar, and Jolma Combined (hg19)</option> + </param> + </inputs> + <outputs> + <data name="out_enriched" format="tabular" label="Enriched Motifs"/> + <data name="out_depleted" format="tabular" label="Depleted Motifs"/> + <data name="out_plots" format="png" label="Motif Count Comparison Plots"/> + </outputs> + + <help> + This tools reads in two counts file and determines enriched and depleted + motifs in two different region sets based on poisson calculation with + gc correction. + </help> + +</tool> \ No newline at end of file
Binary file region-motif-compare/region_motif_db/pouya_test_dir/AP-1_disc1_8mer_C1.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/AP-1_disc1_8mer_C10.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/AP-1_disc1_8mer_C2.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/AP-1_disc1_8mer_C3.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/AP-1_disc1_8mer_C4.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/AP-1_disc1_8mer_C5.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/AP-1_disc1_8mer_C6.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/AP-1_disc1_8mer_C7.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/AP-1_disc1_8mer_C8.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/AP-1_disc1_8mer_C9.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/ZNF263_disc1_8mer_C1.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/ZNF263_disc1_8mer_C2.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/ZNF263_disc1_8mer_C3.RData has changed
Binary file region-motif-compare/region_motif_db/pouya_test_dir/ZNF263_disc1_8mer_C4.RData has changed
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/region_motif_intersect.r Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,102 @@ +# Name: region_motif_intersect.r +# Description: Takes a bed file of target regions and counts intersections +# of each motif (built in rdata database) and target regions. +# Author: Jeremy liu +# Email: jeremy.liu@yale.edu +# Date: 14/07/02 +# Note: This script is meant to be invoked with the following command +# R --slave --vanilla -f ./region_motif_intersect.r --args <workingdir> <db> <inbed> <outtab> +# <workingdir> is working directory of galaxy installation +# <db> types: "t" test, "p" pouya, "j" jaspar jolma, "m" mouse +# Dependencies: none + +# Auxiliary function to concatenate multiple strings +concat <- function(...) { + input_list <- list(...) + return(paste(input_list, sep="", collapse="")) +} + +# Set common and data directories +args <- commandArgs() +workingDir = args[7] +dbDir = concat(workingDir, "/region_motif_db") +dbCode = args[8] +if (dbCode == "t") { + motifDB = concat(dbDir, "/pouya_test_motifs.bed.bgz") +} else if (dbCode == "p") { + motifDB = concat(dbDir, "/pouya_motifs.bed.bgz") +} else if (dbCode == "j") { + motifDB = concat(dbDir, "/jaspar_jolma_motifs.bed.bgz") +} else if (dbCode == "m") { + motifDB = concat(dbDir, "/mm9_motifs.bed.bgz") +} else { + motifDB = concat(dbDir, "/pouya_motifs.bed.bgz") +} + +# Set input and reference files, comment to toggle commmand line arguments +inBed = args[9] +outTab = args[10] + +# Auxiliary function to read in BED file +read_bed <- function(file) { + return(read.table(file, sep="\t", stringsAsFactors=FALSE)) +} + +startTime = Sys.time() +cat("Running ... Started at:", format(startTime, "%a %b %d %X %Y"), "...\n") + +# Load dependencies +cat("Loading dependencies...\n") +suppressPackageStartupMessages(library(Rsamtools, quietly=TRUE)) + +# Initializing hash table (as env) with motif names and loading tabix file +cat("Loading motif database and initializing hash table...\n") +motifTable = new.env() +motifTbx <- TabixFile(motifDB) + +# Loading input bed file, convert integer columns to numeric, name columns +cat("Loading region file...\n") +regionsDF = read_bed(inBed) +dfTemp = sapply(regionsDF, is.integer) +regionsDF[dfTemp] = lapply(regionsDF[dfTemp], as.numeric) +names(regionsDF)[names(regionsDF) == "V1"] = "chr" +names(regionsDF)[names(regionsDF) == "V2"] = "start" +names(regionsDF)[names(regionsDF) == "V3"] = "end" + +# Filtering regions to exclude chromosomes not in motif database +cat("Determining intersection counts...\n") +motifTbxChrs = seqnamesTabix(motifTbx) +regionsDFFilter = subset(regionsDF, chr %in% motifTbxChrs) + +# Loading regions into GRanges object and scanning motif tabix database +# Region end is incremented by 1 since scanTabix querying is inclusive for +# position start but exclusive for position end. +param = GRanges(regionsDFFilter$chr, IRanges(regionsDFFilter$start, + end=regionsDFFilter$end + 1)) +regionsIntersects = scanTabix(motifTbx, param=param) + +# Parsing result list and updating motif count hash table +cat("Parsing result list...\n") +for(regionIntersects in regionsIntersects) { + for(regionIntersect in strsplit(regionIntersects, " ")) { + intersectMotif = strsplit(regionIntersect, "\t")[[1]][4] + if(is.null(motifTable[[intersectMotif]])) { + motifTable[[intersectMotif]] = 1 + } else { + motifTable[[intersectMotif]] = motifTable[[intersectMotif]] + 1 + } + } +} + +# Converting motif count hash table to an integer vector for output +counts = integer(length = length(ls(motifTable))) +names(counts) = ls(motifTable) +for(motifName in ls(motifTable)) { + counts[motifName] = as.integer(motifTable[[motifName]]) +} + +# Outputting intersection counts to tab delineated file +cat("Outputting to file...\n") +write.table(counts, outTab, quote=FALSE, sep="\t", row.names=TRUE, col.names=FALSE) +cat("Done. Job started at:", format(startTime, "%a %b %d %X %Y."), + "Job ended at:", format(Sys.time(), "%a %b %d %X %Y."), "\n")
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/region_motif_intersect.xml Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,27 @@ +<tool id="region_motif_intersect" name="Region Motif Intersect"> + <description>for computing the motifs that lie inside a region set</description> + <requirements> + <requirement type="set_environment">RMOTIF_PATH</requirement> + </requirements> + <command interpreter="Rscript"> + region_motif_intersect.r --args \$RMOTIF_PATH $db_type $in_bed $out_tab + </command> + <inputs> + <param name="in_bed" type="data" format="bed" label="Input BED File" /> + <param name="db_type" type="select" label="Select Motif Database" > + <option value="t">Test Pouya Subset (hg19)</option> + <option value="p">Pouya Encode Motifs (hg19)</option> + <option value="j">Jaspar and Jolma Motifs (hg19)</option> + <option value="m">Mouse Motifs (mm9)</option> + </param> + </inputs> + <outputs> + <data name="out_tab" format="tabular" /> + </outputs> + + <help> + This tool computes the motifs and the number of motifs that intersect + any region in a input set of regions. + </help> + +</tool> \ No newline at end of file
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/region_motif_lib/plotting.r Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,815 @@ +library(graphics, quietly=TRUE) + +plot.verbose=F +name.cleaner<-function(...,sep="",replace="_") { + plot.name=gsub(" ",replace,paste(...,sep=sep)) + plot.name=gsub("/",replace,plot.name) + plot.name=gsub(",",replace,plot.name) + plot.name=gsub("'",replace,plot.name) + plot.name=gsub("\\+","plus",plot.name) + plot.name=gsub("\\(","",plot.name) + plot.name=gsub("\\)","",plot.name) + return(plot.name) +} +plot.namer <- function(..., date=0, fig.dir=0, format="png",sep="") { + plot.name=name.cleaner(...,sep=sep) + if(date==0) date=gsub("-","",as.character(Sys.Date())) + if(fig.dir==0) fig.dir="/Users/alver/allplots" + plot.name=paste(fig.dir,"/",date,plot.name,".",format,sep="") + if(plot.verbose) cat(" saving figure: ",plot.name,"\n") + return(plot.name) +} + +plot.scatter <- function(x,y=NULL,f=0.1,same=FALSE,n.points=-1,draw.lowess=FALSE,write.r=TRUE,cex.r=1,col=NULL,col.line=NULL,lwd.line=1, + draw.loess=FALSE,span=0.5,bandwidth=bandwidth,draw.prof=FALSE,xlog=FALSE,ylog=FALSE,cor.method="pearson",log="",ind=NULL, + draw.spread=FALSE,...) { + + ## if col is the same length as x, use col for each point matching x. + ## if col is the same length as ind, use col for each point matching x[ind]. + ## else use densCols function based on col. + ## if col is null, densCols is used with bluetone for first plot and redtone for same=T. + + #print(length(x)) + #print(length(y)) + + xy <- xy.coords(x, y) + x=xy$x + y=xy$y + + output=list() + col.use = col + + if(!is.null(ind)) { + if(length(col.use)==length(x)) { + col.use=col.use[ind] + } + x=x[ind] + y=y[ind] + } + + if(length(col.use)!=length(x)) { + col.use=rep(NA,length(x)) + } + + + take=which(is.finite(x) & is.finite(y)) + x=x[take] + y=y[take] + col.use=col.use[take] + + if(grepl("x",log)) xlog=TRUE + if(grepl("y",log)) ylog=TRUE + if(xlog) log="x" + if(ylog) log=paste(log,"y",sep="") + + if(draw.lowess) { + lo = lowess(x,y,f) + output$lowess=lo + } + if(draw.loess | draw.spread) { + px=x;py=y + if(xlog) px=log(x) + if(ylog) py=log(y) + ind = which(is.finite(px+py)) + px=px[ind] + py=py[ind] + lo = loess(py ~ px,span=span,iterations=5) + lo.y=as.numeric(lo$fitted) + lo.x=as.numeric(lo$x) + if(draw.spread) lo.sd = loess((lo.y-py)^2 ~ lo.x,span=span*1.5,iterations=5) + if(xlog) lo.x=exp(lo.x) + if(ylog) lo.y=exp(lo.y) + lo =data.frame(x=lo.x,y=lo.y) + if(draw.spread) { + lo.sd=lo.sd$fitted + if(ylog) lo.sd=lo.sd*lo.y*lo.y + lo$sd=sqrt(pmax(0,lo.sd)) + } + lo=unique(lo) + lo = lo[order(lo$x),] + output$loess=lo + } + + if(draw.prof) { + px=x;py=y + if(xlog) px=log(x) + p=prof(px,py,50) + if(xlog) p$x=exp(p$x) + output$prof=p + } + + r=cor(x,y,method=cor.method) + output$cor=r + output$cor.method=cor.method + + len=length(x) + if(n.points>0 & n.points<len) { + take=sample(1:len,n.points) + x=x[take] + y=y[take] + col.use=col.use[take] + } + + if(xlog) { + ind = which(x>0) + x=x[ind] + y=y[ind] + col.use=col.use[ind] + } + xcol=x + if(xlog) xcol=log(xcol) + if(ylog) { + ind = which(y>0) + x=x[ind] + xcol=xcol[ind] + y=y[ind] + col.use=col.use[ind] + } + ycol=y + if(ylog) ycol=log(ycol) + + if(is.null(col)) { + if(!same) { + col=colorRampPalette(blues9[-(1:3)]) + } else { + col=colorRampPalette(c("lightpink","red","darkred")) + } + } + if(!is.na(col.use[1])) { + col=col.use + } else { + col= suppressPackageStartupMessages(densCols(xcol,ycol,col =col,bandwidth=bandwidth,nbin=500)) + } + if(!same) { + plot(x,y,col=col,log=log,...) + } else { + points(x,y,col=col,...) + } + + if(is.null(col.line)) { + col.line="darkblue" + if(same) col.line="darkred" + } + if(draw.lowess | draw.loess) lines(lo,col=col.line,lwd=lwd.line) + if(draw.spread) { + lines(lo$x,lo$y+lo$sd,col=col.line,lwd=lwd.line) + lines(lo$x,lo$y-lo$sd,col=col.line,lwd=lwd.line) + } + if(draw.prof) { + points(p) + plot.prof(p) + } + if(write.r & !same) mtext(paste("r=",round(r,3),sep=""),cex=cex.r) + return(invisible(output)) +} + +#color.int=c(144,586,465,257,490,100,74,24) +#coli=1 +#cols = integer() +colramp.bwr = vector() +colramp.byr = vector() +colramp.bw = vector() +colramp.bw2 = vector() + +plot.save=F + +setup.plotting <- function() { + pdf.options(useDingbats = FALSE) +# cols<<-colors()[color.int] +# cols<<-rep(cols,100) + colramp.bwr <<- colorRampPalette(c("blue","white","red"),space="Lab")(100); + colramp.byr <<- colorRampPalette(c("blue","yellow","red"),space="Lab")(100); + colramp.bw <<- colorRampPalette(c("white","black"),space="Lab")(100) + colramp.bw2 <<- colorRampPalette(c("grey92","grey5"),space="Lab")(100) +} + + +plot.cluster <- function(x,k, max.points.cl=-1, image.sep=-1,col=NULL, reorder=FALSE) { + x[which(is.na(x))]=0 + if(reorder) { + o=hclust(dist(t(x)))$order + x=x[,o] + } + if(image.sep<0) { + if(max.points.cl>0) { + image.sep=ceiling(0.2*max.points.cl) + } else { + image.sep=ceiling(0.2 * nrow(x) / nrow(k$centers)) + } + } + + distances<-dist(k$centers) + hcl=hclust(distances) + + adjust.branch.sep <-function(ddr,lengths) { + assign.branch.sep <- function(d,i.leaf) { + if(is.leaf(d)) { + attr(d,"members")<-lengths[i.leaf] + i.leaf=i.leaf+1 + output=list(d=d,i.leaf=i.leaf) + return(output) + } + else{ + input=assign.branch.sep(d[[1]],i.leaf) + i.leaf=input$i.leaf + d[[1]]=input$d + + input=assign.branch.sep(d[[2]],i.leaf) + i.leaf=input$i.leaf + d[[2]]=input$d + + attr(d,"members")<-attr(d[[1]],"members")+attr(d[[2]],"members") + output=list(d=d,i.leaf=i.leaf) + return(output) + } + } + ddr<-as.dendrogram(ddr) + ddr=assign.branch.sep(ddr,1)$d + return(ddr) + } + + n.points.actual=k$size + if(max.points.cl>0) { + k$size[which(k$size>max.points.cl)] = max.points.cl +} + + ddr<-adjust.branch.sep(hcl,k$size[hcl$order]+image.sep) + centers=length(hcl$order) + + n.points=sum(k$size) + n.dims=ncol(x) + z=matrix(numeric((n.points+(centers-1)*image.sep)*n.dims),ncol=n.dims) + + + last.row=0 + cluster.y.pos=numeric(centers) + for(i.c in hcl$order) { + n.p=k$size[i.c] + z[last.row+1:n.p,] = x[which(k$cluster==i.c)[1:n.p],] + cluster.y.pos[i.c]=last.row+n.p/2 + last.row=last.row+n.p+image.sep + } + + zlim=c(0,max(z)) + if(min(z)<0) { + m=max(c(z,-z)) + zlim=c(-m,m) + } + if(is.null(col)) { + if(min(z)>=0) { + col= colramp.bw + } else { + col= colorRampPalette(c("blue","yellow","red"),space="Lab")(100); + } + } + x.pl=seq1(n.dims+1)-0.5 + y.pl=seq1(nrow(z)+1)-0.5 + l <- layout(matrix(1:2,ncol=2),widths=c(1,5)) + par(mar = c(6,0.5,6,0)) + my.plot.dendrogram(ddr,horiz=T,axes=F,yaxs="i",xaxs="i",leaflab="none",center=T,lwd=10) + par(mar = c(6,0.1,6,2.1)) + image(x=x.pl,y=y.pl,z=t(z),zlim=zlim,axes=FALSE,xlab="",col=col) + mtext("cluster",side=4,adj=1.1) + mtext("points",side=4,adj=1.1,line=1) + mtext(seq1(centers),side=4,at=cluster.y.pos) + mtext(n.points.actual,side=4,at=cluster.y.pos,line=1) + + if(!is.null(dimnames(x)[[2]])) { + mtext(dimnames(x)[[2]],side=1,at=seq1(n.dims),las=2) + } +} + +plot.cluster2 <- function(k, n.clusters=-1, n.clusters.per.panel=4, cols=c("black","red","blue","darkgreen","orange"),f=0,xshift=0,...) { + if(n.clusters<=0) n.clusters=nrow(k$centers) + + n.elements=as.numeric(unlist(lapply(seq1(n.clusters), function(cl) length(which(abs(k$cluster)==cl))))) + + distances<-dist(k$centers) + n.panels = ceiling(n.clusters/n.clusters.per.panel) + n.rows=ceiling(sqrt(n.panels)) + n.cols=ceiling(n.panels/n.rows) + n.panels.layout=n.rows*n.cols + + layout(matrix(seq1(n.panels.layout),nrow=n.rows,byrow=TRUE)) + + min=min(k$centers) + max=max(k$centers) + + if(f>0) { + for(i.cluster in seq1(n.clusters)) { + k$centers[i.cluster,]=lowess(k$centers[i.cluster,],f=f)$y + } + } + + ## hcl=hclust(distances) + hcl=list() + hcl$order=1:n.clusters + + for(i.cluster in seq1(n.clusters)) { + if(i.cluster %% n.clusters.per.panel == 1 ) { + clusters.of.panel=i.cluster:(i.cluster+n.clusters.per.panel-1) + clusters.of.panel=clusters.of.panel[which(clusters.of.panel<=n.clusters)] + clusters.of.panel=hcl$order[clusters.of.panel] + plot(c(0,length(k$centers[1,]))+xshift,c(min,max),type="n",...) + mtext(paste(clusters.of.panel," (",n.elements[clusters.of.panel],")",sep=""),line=length(clusters.of.panel)-seq1(length(clusters.of.panel)),col=cols[seq1(length(clusters.of.panel)) %% n.clusters.per.panel+1] ) + } + # lines(k$centers[hcl$order[i.cluster],],col=cols[i.cluster %% n.clusters.per.panel+1]) + lines(seq1(length(k$centers[1,]))+xshift,k$centers[hcl$order[i.cluster],],col=cols[i.cluster %% n.clusters.per.panel+1]) + } +} + +my.colors <- function(n) { + few.colors=c("black","red","blue","green3","mediumorchid3","gold2","darkcyan","sienna2") + if(n<=length(few.colors)) return(few.colors [seq1(n)]) + col=integer(n) + n.families=7 + n.members=ceiling(n/n.families) + for(i in seq1(n)) { + member=ceiling(i/n.families) + ratio=(member-1)/(n.members-1) + c2=0+0.8*ratio + if(member %% 2 == 1) ratio=-ratio + c1=0.8-0.2*ratio + c3=0.75-0.2*ratio + if(i %% n.families == 1) {col[i]=rgb(c2,c2,c2)} + if(i %% n.families == 2) {col[i]=rgb(c1,c1/2,c1/2)} + if(i %% n.families == 3) {col[i]=rgb(c1/2,0.9*c1,c1/2)} + if(i %% n.families == 4) {col[i]=rgb(c1/2,c1/2,c1)} + if(i %% n.families == 5) {col[i]=rgb(c3,c3,c3/2)} + if(i %% n.families == 6) {col[i]=rgb(c3,c3/2,c3)} + if(i %% n.families == 0) {col[i]=rgb(c3/2,c3,c3)} + } + return(col) +} + +plot.my.colors <-function(n) { + x11() + col=my.colors(n) + plot(x=c(0,n),y=c(0,1),type="n") + segments(seq1(n)-1,runif(n),seq1(n),runif(n),col=col) +} + + +plot.colors <-function() { + x11(width=10,height=10) + plot(c(0,26),c(0,26),type="n") + c=colors() + n=length(c) + i=1:n + x=i%%26 + y=floor(i/26) + rect(x,y,x+1,y+1,col=c[i],border=c[i]) + text(x+0.5,y+0.5,i) +} + + +adjust.branch.sep <-function(ddr,lengths) { + assign.branch.sep <- function(d,i.leaf) { + if(is.leaf(d)) { + attr(d,"members")<-lengths[i.leaf] + i.leaf=i.leaf+1 + output=list(d=d,i.leaf=i.leaf) + return(output) + } + else{ + input=assign.branch.sep(d[[1]],i.leaf) + i.leaf=input$i.leaf + d[[1]]=input$d + + input=assign.branch.sep(d[[2]],i.leaf) + i.leaf=input$i.leaf + d[[2]]=input$d + + attr(d,"members")<-attr(d[[1]],"members")+attr(d[[2]],"members") + output=list(d=d,i.leaf=i.leaf) + return(output) + } + } + ddr<-as.dendrogram(ddr) + ddr=assign.branch.sep(ddr,1)$d + return(ddr) +} +t.dhcol <- function(dr,h,cols=c(1)) { + # check child heights + if(attr(dr[[1]],"height")<h) { + # color + ecol <- cols[coli]; + coli <<- coli+1; + dr[[1]] <- dendrapply(dr[[1]],function(e) { attr(e,"edgePar") <- list(col=ecol); e}); + attr(dr[[1]],"edgePar") <- list(col=ecol,p.border=NA,p.col=NA,t.col=1,t.cex=1.3); + } else { + dr[[1]] <- t.dhcol(dr[[1]],h,cols); + } + + if(attr(dr[[2]],"height")<h) { + # color + ecol <- cols[coli]; + coli <<- coli+1; + dr[[2]] <- dendrapply(dr[[2]],function(e) { attr(e,"edgePar") <- list(col=ecol); e}); + attr(dr[[2]],"edgePar") <- list(col=ecol,p.border=NA,p.col=NA,t.col=1,t.cex=1.3); + } else { + dr[[2]] <- t.dhcol(dr[[2]],h,cols); + } + return(dr); +} + + + +### The rest is PeterK's my.plot.dendogram + +## FIXME: need larger par("mar")[1] or [4] for longish labels ! +## {probably don't change, just print a warning ..} +my.plot.dendrogram <- + function (x, type = c("rectangle", "triangle"), center = FALSE, + edge.root = is.leaf(x) || !is.null(attr(x, "edgetext")), + nodePar = NULL, edgePar = list(), + leaflab = c("perpendicular", "textlike", "none"), dLeaf = NULL, + xlab = "", ylab = "", xaxt="n", yaxt="s", + horiz = FALSE, frame.plot = FALSE, ...) +{ + type <- match.arg(type) + leaflab <- match.arg(leaflab) + hgt <- attr(x, "height") + if (edge.root && is.logical(edge.root)) + edge.root <- 0.0625 * if(is.leaf(x)) 1 else hgt + mem.x <- .my.memberDend(x) + yTop <- hgt + edge.root + if(center) { x1 <- 0.5 ; x2 <- mem.x + 0.5 } + else { x1 <- 1 ; x2 <- mem.x } + xlim <- c(x1 - 1/2, x2 + 1/2) + ylim <- c(0, yTop) + if (horiz) {## swap and reverse direction on `x': + xl <- xlim; xlim <- rev(ylim); ylim <- xl + tmp <- xaxt; xaxt <- yaxt; yaxt <- tmp + } + plot(0, xlim = xlim, ylim = ylim, type = "n", xlab = xlab, ylab = ylab, + xaxt = xaxt, yaxt = yaxt, frame.plot = frame.plot, ...) + if(is.null(dLeaf)) + dLeaf <- .75*(if(horiz) strwidth("w") else strheight("x")) + + if (edge.root) { +### FIXME: the first edge + edgetext is drawn here, all others in plotNode() +### ----- maybe use trick with adding a single parent node to the top ? + x0 <- my.plotNodeLimit(x1, x2, x, center)$x + if (horiz) + segments(hgt, x0, yTop, x0) + else segments(x0, hgt, x0, yTop) + if (!is.null(et <- attr(x, "edgetext"))) { + my <- mean(hgt, yTop) + if (horiz) + text(my, x0, et) + else text(x0, my, et) + } + } + my.plotNode(x1, x2, x, type = type, center = center, leaflab = leaflab, + dLeaf = dLeaf, nodePar = nodePar, edgePar = edgePar, horiz = horiz) +} + +### the work horse: plot node (if pch) and lines to all children +my.plotNode <- + function(x1, x2, subtree, type, center, leaflab, dLeaf, + nodePar, edgePar, horiz = FALSE) +{ + inner <- !is.leaf(subtree) && x1 != x2 + yTop <- attr(subtree, "height") + bx <- my.plotNodeLimit(x1, x2, subtree, center) + xTop <- bx$x + usrpar <- par("usr"); + + ## handle node specific parameters in "nodePar": + hasP <- !is.null(nPar <- attr(subtree, "nodePar")) + if(!hasP) nPar <- nodePar + + if(getOption("verbose")) { + cat(if(inner)"inner node" else "leaf", ":") + if(!is.null(nPar)) { cat(" with node pars\n"); str(nPar) } + cat(if(inner)paste(" height", formatC(yTop),"; "), + "(x1,x2)= (",formatC(x1,wid=4),",",formatC(x2,wid=4),")", + "--> xTop=", formatC(xTop, wid=8),"\n", sep="") + } + + Xtract <- function(nam, L, default, indx) + rep(if(nam %in% names(L)) L[[nam]] else default, + length.out = indx)[indx] + asTxt <- function(x) # to allow 'plotmath' labels: + if(is.character(x) || is.expression(x) || is.null(x)) x else as.character(x) + + i <- if(inner || hasP) 1 else 2 # only 1 node specific par + + if(!is.null(nPar)) { ## draw this node + pch <- Xtract("pch", nPar, default = 1:2, i) + cex <- Xtract("cex", nPar, default = c(1,1), i) + col <- Xtract("col", nPar, default = par("col"), i) + bg <- Xtract("bg", nPar, default = par("bg"), i) + points(if (horiz) cbind(yTop, xTop) else cbind(xTop, yTop), + pch = pch, bg = bg, col = col, cex = cex) + } + + if(leaflab == "textlike") + p.col <- Xtract("p.col", nPar, default = "white", i) + lab.col <- Xtract("lab.col", nPar, default = par("col"), i) + lab.cex <- Xtract("lab.cex", nPar, default = c(1,1), i) + lab.font <- Xtract("lab.font", nPar, default = par("font"), i) + if (is.leaf(subtree)) { + ## label leaf + if (leaflab == "perpendicular") { # somewhat like plot.hclust + if(horiz) { + X <- yTop + dLeaf * lab.cex + Y <- xTop; srt <- 0; adj <- c(0, 0.5) + } + else { + Y <- yTop - dLeaf * lab.cex + X <- xTop; srt <- 90; adj <- 1 + } + nodeText <- asTxt(attr(subtree,"label")) + text(X, Y, nodeText, xpd = TRUE, srt = srt, adj = adj, + cex = lab.cex, col = lab.col, font = lab.font) + } + } + else if (inner) { + segmentsHV <- function(x0, y0, x1, y1) { + if (horiz) + segments(y0, x0, y1, x1, col = col, lty = lty, lwd = lwd) + else segments(x0, y0, x1, y1, col = col, lty = lty, lwd = lwd) + } + for (k in 1:length(subtree)) { + child <- subtree[[k]] + ## draw lines to the children and draw them recursively + yBot <- attr(child, "height") + if (getOption("verbose")) cat("ch.", k, "@ h=", yBot, "; ") + if (is.null(yBot)) + yBot <- 0 + xBot <- + if (center) mean(bx$limit[k:(k + 1)]) + else bx$limit[k] + .my.midDend(child) + + hasE <- !is.null(ePar <- attr(child, "edgePar")) + if (!hasE) + ePar <- edgePar + i <- if (!is.leaf(child) || hasE) 1 else 2 + ## define line attributes for segmentsHV(): + col <- Xtract("col", ePar, default = par("col"), i) + lty <- Xtract("lty", ePar, default = par("lty"), i) + lwd <- Xtract("lwd", ePar, default = par("lwd"), i) + if (type == "triangle") { + segmentsHV(xTop, yTop, xBot, yBot) + } + else { # rectangle + segmentsHV(xTop,yTop, xBot,yTop)# h + segmentsHV(xBot,yTop, xBot,yBot)# v + } + vln <- NULL + if (is.leaf(child) && leaflab == "textlike") { + nodeText <- asTxt(attr(child,"label")) + if(getOption("verbose")) + cat('-- with "label"',format(nodeText)) + hln <- 0.6 * strwidth(nodeText, cex = lab.cex)/2 + vln <- 1.5 * strheight(nodeText, cex = lab.cex)/2 + rect(xBot - hln, yBot, + xBot + hln, yBot + 2 * vln, col = p.col) + text(xBot, yBot + vln, nodeText, xpd = TRUE, + cex = lab.cex, col = lab.col, font = lab.font) + } + if (!is.null(attr(child, "edgetext"))) { + edgeText <- asTxt(attr(child, "edgetext")) + if(getOption("verbose")) + cat('-- with "edgetext"',format(edgeText)) + if (!is.null(vln)) { + mx <- + if(type == "triangle") + (xTop+ xBot+ ((xTop - xBot)/(yTop - yBot)) * vln)/2 + else xBot + my <- (yTop + yBot + 2 * vln)/2 + } + else { + mx <- if(type == "triangle") (xTop + xBot)/2 else xBot + my <- (yTop + yBot)/2 + } + ## Both for "triangle" and "rectangle" : Diamond + Text + + p.col <- Xtract("p.col", ePar, default = "white", i) + p.border <- Xtract("p.border", ePar, default = par("fg"), i) + ## edge label pars: defaults from the segments pars + p.lwd <- Xtract("p.lwd", ePar, default = lwd, i) + p.lty <- Xtract("p.lty", ePar, default = lty, i) + t.col <- Xtract("t.col", ePar, default = col, i) + t.cex <- Xtract("t.cex", ePar, default = 1, i) + t.font<- Xtract("t.font",ePar, default= par("font"), i) + t.shift <- Xtract("t.shift", ePar, default = 0.01, i) + + vlm <- strheight(c(edgeText,"h"), cex = t.cex)/2 + hlm <- strwidth (c(edgeText,"m"), cex = t.cex)/2 + hl3 <- c(hlm[1], hlm[1] + hlm[2], hlm[1]) + #polygon(mx+ c(-hl3, hl3), my + sum(vlm)*c(-1:1,1:-1), + # col = p.col, border= p.border, lty = p.lty, lwd = p.lwd) + #text(mx, my, edgeText, cex = t.cex, col = t.col, font = t.font) + if(horiz) { + text(my, mx+t.shift*abs(usrpar[3]-usrpar[4]), edgeText, cex = t.cex, col = t.col, font = t.font) + } else { + text(mx+t.shift*abs(usrpar[2]-usrpar[1]), my, edgeText, cex = t.cex, col = t.col, font = t.font) + } + } + my.plotNode(bx$limit[k], bx$limit[k + 1], subtree = child, + type, center, leaflab, dLeaf, nodePar, edgePar, horiz) + } + } +} + +my.plotNodeLimit <- function(x1, x2, subtree, center) +{ + ## get the left borders limit[k] of all children k=1..K, and + ## the handle point `x' for the edge connecting to the parent. + inner <- !is.leaf(subtree) && x1 != x2 + if(inner) { + K <- length(subtree) + mTop <- .my.memberDend(subtree) + limit <- integer(K) + xx1 <- x1 + for(k in 1:K) { + m <- .my.memberDend(subtree[[k]]) + ##if(is.null(m)) m <- 1 + xx1 <- xx1 + (if(center) (x2-x1) * m/mTop else m) + limit[k] <- xx1 + } + limit <- c(x1, limit) + } else { ## leaf + limit <- c(x1, x2) + } + mid <- attr(subtree, "midpoint") + center <- center || (inner && !is.numeric(mid)) + x <- if(center) mean(c(x1,x2)) else x1 + (if(inner) mid else 0) + list(x = x, limit = limit) +} + +.my.memberDend <- function(x) { + r <- attr(x,"x.member") + if(is.null(r)) { + r <- attr(x,"members") + if(is.null(r)) r <- 1:1 + } + r +} + +.my.midDend <- function(x) + if(is.null(mp <- attr(x, "midpoint"))) 0 else mp + + +## original Andy Liaw; modified RG, MM : +my.heatmap <- function (x, Rowv=NULL, Colv=if(symm)"Rowv" else NULL, + distfun = dist, hclustfun = hclust, + reorderfun = function(d,w) reorder(d,w), + add.expr, symm = FALSE, revC = identical(Colv, "Rowv"), + scale = c("row", "column", "none"), na.rm=TRUE, + margins = c(5, 5), ColSideColors, RowSideColors, + cexRow = 0.2 + 1/log10(nr), cexCol = 0.2 + 1/log10(nc), + labRow = NULL, labCol = NULL, main = NULL, xlab = NULL, ylab = NULL, + keep.dendro = FALSE, + verbose = getOption("verbose"), imageSize=4, imageVSize=imageSize,imageHSize=imageSize,lasCol=2, lasRow=2, respect=F, ...) +{ + scale <- if(symm && missing(scale)) "none" else match.arg(scale) + if(length(di <- dim(x)) != 2 || !is.numeric(x)) + stop("'x' must be a numeric matrix") + nr <- di[1] + nc <- di[2] + if(nr <= 1 || nc <= 1) + stop("'x' must have at least 2 rows and 2 columns") + if(!is.numeric(margins) || length(margins) != 2) + stop("'margins' must be a numeric vector of length 2") + + doRdend <- !identical(Rowv,NA) + doCdend <- !identical(Colv,NA) + ## by default order by row/col means + if(is.null(Rowv)) Rowv <- rowMeans(x, na.rm = na.rm) + if(is.null(Colv)) Colv <- colMeans(x, na.rm = na.rm) + + ## get the dendrograms and reordering indices + + if(doRdend) { + if(inherits(Rowv, "dendrogram")) + ddr <- Rowv + else { + hcr <- hclustfun(distfun(x)) + ddr <- as.dendrogram(hcr) + if(!is.logical(Rowv) || Rowv) + ddr <- reorderfun(ddr, Rowv) + } + if(nr != length(rowInd <- order.dendrogram(ddr))) + stop("row dendrogram ordering gave index of wrong length") + } + else rowInd <- 1:nr + + if(doCdend) { + if(inherits(Colv, "dendrogram")) + ddc <- Colv + else if(identical(Colv, "Rowv")) { + if(nr != nc) + stop('Colv = "Rowv" but nrow(x) != ncol(x)') + ddc <- ddr + } + else { + hcc <- hclustfun(distfun(if(symm)x else t(x))) + ddc <- as.dendrogram(hcc) + if(!is.logical(Colv) || Colv) + ddc <- reorderfun(ddc, Colv) + } + if(nc != length(colInd <- order.dendrogram(ddc))) + stop("column dendrogram ordering gave index of wrong length") + } + else colInd <- 1:nc + + ## reorder x + x <- x[rowInd, colInd] + + labRow <- + if(is.null(labRow)) + if(is.null(rownames(x))) (1:nr)[rowInd] else rownames(x) + else labRow[rowInd] + labCol <- + if(is.null(labCol)) + if(is.null(colnames(x))) (1:nc)[colInd] else colnames(x) + else labCol[colInd] + + if(scale == "row") { + x <- sweep(x, 1, rowMeans(x, na.rm = na.rm)) + sx <- apply(x, 1, sd, na.rm = na.rm) + x <- sweep(x, 1, sx, "/") + } + else if(scale == "column") { + x <- sweep(x, 2, colMeans(x, na.rm = na.rm)) + sx <- apply(x, 2, sd, na.rm = na.rm) + x <- sweep(x, 2, sx, "/") + } + + ## Calculate the plot layout + lmat <- rbind(c(NA, 3), 2:1) + lwid <- c(if(doRdend) 1 else 0.05, imageHSize) + lhei <- c((if(doCdend) 1 else 0.05) + if(!is.null(main)) 0.2 else 0, imageVSize) + if(!missing(ColSideColors)) { ## add middle row to layout + if(!is.character(ColSideColors) || length(ColSideColors) != nc) + stop("'ColSideColors' must be a character vector of length ncol(x)") + lmat <- rbind(lmat[1,]+1, c(NA,1), lmat[2,]+1) + lhei <- c(lhei[1], 0.2, lhei[2]) + } + if(!missing(RowSideColors)) { ## add middle column to layout + if(!is.character(RowSideColors) || length(RowSideColors) != nr) + stop("'RowSideColors' must be a character vector of length nrow(x)") + lmat <- cbind(lmat[,1]+1, c(rep(NA, nrow(lmat)-1), 1), lmat[,2]+1) + lwid <- c(lwid[1], 0.2, lwid[2]) + } + lmat[is.na(lmat)] <- 0 + if(verbose) { + cat("layout: widths = ", lwid, ", heights = ", lhei,"; lmat=\n") + print(lmat) + } + + ## Graphics `output' ----------------------- + + op <- par(no.readonly = TRUE) + on.exit(par(op)) + layout(lmat, widths = lwid, heights = lhei, respect = respect) + ## draw the side bars + if(!missing(RowSideColors)) { + par(mar = c(margins[1],0, 0,0.5)) + image(rbind(1:nr), col = RowSideColors[rowInd], axes = FALSE) + } + if(!missing(ColSideColors)) { + par(mar = c(0.5,0, 0,margins[2])) + image(cbind(1:nc), col = ColSideColors[colInd], axes = FALSE) + } + ## draw the main carpet + par(mar = c(margins[1], 0, 0, margins[2])) + if(!symm || scale != "none") + x <- t(x) + if(revC) { # x columns reversed + iy <- nr:1 + ddr <- rev(ddr) + x <- x[,iy] + } else iy <- 1:nr + + image(1:nc, 1:nr, x, xlim = 0.5+ c(0, nc), ylim = 0.5+ c(0, nr), + axes = FALSE, xlab = "", ylab = "", ...) + axis(1, 1:nc, labels= labCol, las= lasCol, line= -0.5, tick= 0, cex.axis= cexCol) + if(!is.null(xlab)) mtext(xlab, side = 1, line = margins[1] - 1.25) + axis(4, iy, labels= labRow, las= lasRow, line= -0.5, tick= 0, cex.axis= cexRow) + if(!is.null(ylab)) mtext(ylab, side = 4, line = margins[2] - 1.25,las=lasRow) + if (!missing(add.expr)) + eval(substitute(add.expr)) + + ## the two dendrograms : + par(mar = c(margins[1], 0, 0, 0)) + if(doRdend) + my.plot.dendrogram(ddr, horiz = TRUE, axes = FALSE, yaxs = "i", leaflab = "none") + else frame() + + par(mar = c(0, 0, if(!is.null(main)) 1 else 0, margins[2])) + if(doCdend) + my.plot.dendrogram(ddc, axes = FALSE, xaxs = "i", leaflab = "none") + else if(!is.null(main)) frame() + + ## title + if(!is.null(main)) title(main, cex.main = 1.5*op[["cex.main"]]) + + invisible(list(rowInd = rowInd, colInd = colInd, + Rowv = if(keep.dendro && doRdend) ddr, + Colv = if(keep.dendro && doCdend) ddc )) +}
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/region-motif-compare/tool_dependencies.xml Wed Nov 12 15:21:11 2014 -0500 @@ -0,0 +1,6 @@ +<?xml version="1.0"?> +<tool_dependency> + <set_environment version="1.0"> + <environment_variable name="RMOTIF_PATH" action="set_to">$REPOSITORY_INSTALL_DIR</environment_variable> + </set_environment> +</tool_dependency> \ No newline at end of file
--- a/upload/region_motif_compare.r Wed Nov 12 15:19:37 2014 -0500 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,188 +0,0 @@ -# Name: region_motif_compare.r -# Description: Reads in two count files and determines enriched and depleted -# motifs (or any location based feature) based on poisson tests and gc -# corrections. All enrichment ratios relative to overall count / gc ratios. -# Author: Jeremy liu -# Email: jeremy.liu@yale.edu -# Date: 14/07/03 -# Note: This script is meant to be invoked with the following command -# R --slave --vanilla -f ./region_motif_compare.r --args <workingdir> <db> <intab1> <intab2> -# <enriched_tab> <depleted_tab> <plots_png> -# <workingdir> is working directory of galaxy installation -# <db> types: "t" test, "p" pouya, "j" jaspar jolma, "m" mouse, "c" combined -# Dependencies: plotting.r - -# Auxiliary function to concatenate multiple strings -concat <- function(...) { - input_list <- list(...) - return(paste(input_list, sep="", collapse="")) -} - -# Supress all warning messages to prevent Galaxy treating warnings as errors -options(warn=-1) - -# Set common and data directories -args <- commandArgs() -workingDir = args[7] -dbDir = concat(workingDir, "/region_motif_db") -dbCode = args[8] -# dbCode "c" implemented when pwmFile is loaded -if (dbCode == "t" | dbCode == "p") { - pwmFile = concat(dbDir, "/pwms/pouya.pwms.from.seq.RData") -} else if (dbCode == "j") { - pwmFile = concat(dbDir, "/pwms/jaspar.jolma.pwms.from.seq.RData") -} else if (dbCode == "m") { - pwmFile = concat(dbDir, "/pwms/mm9.pwms.from.seq.RData") -} else if (dbCode == "c") { # rest of dbCode "c" implemeted when pwmFile loaded - pwmFile = concat(dbDir, "/pwms/pouya.pwms.from.seq.RData") - pwmFile2 = concat(dbDir, "/pwms/jaspar.jolma.pwms.from.seq.RData") -} else { - pwmFile = concat(dbDir, "/pwms/pouya.pwms.from.seq.RData") -} - -# Set input and reference files -inTab1 = args[9] -inTab2 = args[10] -enrichTab = args[11] -depleteTab = args[12] -plotsPng = args[13] - -# Load dependencies -source(concat(workingDir, "/region_motif_lib/plotting.r")) - -# Auxiliary function to read in tab file and prepare the data -read_tsv <- function(file) { - data = read.table(file, sep="\t", stringsAsFactors=FALSE) - names(data)[names(data) == "V1"] = "motif" - names(data)[names(data) == "V2"] = "counts" - return(data) -} - -startTime = Sys.time() -cat("Running ... Started at:", format(startTime, "%a %b %d %X %Y"), "...\n") - -# Loading motif position weight matrix (pwm) file and input tab file -cat("Loading and reading input region motif count files...\n") -load(pwmFile) # pwms data structure -if (dbCode == "c") { # Remaining implementation of dbCode "c" combined - temp = pwms - load(pwmFile2) - pwms = append(temp, pwms) -} -region1DF = read_tsv(inTab1) -region2DF = read_tsv(inTab2) -region1Counts = region1DF$counts -region2Counts = region2DF$counts -names(region1Counts) = region1DF$motif -names(region2Counts) = region2DF$motif - -# Processing count vectors to account for missing 0 count motifs, then sorting -cat("Performing 0 count correction and sorting...\n") -allNames = union(names(region1Counts), names(region2Counts)) -region1Diff = setdiff(allNames, names(region1Counts)) -region2Diff = setdiff(allNames, names(region2Counts)) -addCounts1 = rep(0, length(region1Diff)) -addCounts2 = rep(0, length(region2Diff)) -names(addCounts1) = region1Diff -names(addCounts2) = region2Diff -newCounts1 = append(region1Counts, addCounts1) -newCounts2 = append(region2Counts, addCounts2) -region1Counts = newCounts1[sort.int(names(newCounts1), index.return=TRUE)$ix] -region2Counts = newCounts2[sort.int(names(newCounts2), index.return=TRUE)$ix] - -# Generate gc content matrix -gc = sapply(pwms, function(i) mean(i[2:3,3:18])) - -# Apply poisson test, calculate p and q values, and filter significant results -cat("Applying poisson test...\n") -rValue = sum(region2Counts) / sum(region1Counts) -pValue = sapply(seq(along=region1Counts), function(i) { - poisson.test(c(region1Counts[i], region2Counts[i]), r=1/rValue)$p.value -}) -qValue = p.adjust(pValue, "fdr") -indices = which(qValue<0.1 & abs(log2(region1Counts/region2Counts/rValue))>log2(1.5)) - -# Setting up output diagnostic plots, 4 in 1 png image -png(plotsPng, width=800, height=800) -xlab = "region1_count" -ylab = "region2_count" -lim = c(0.5, 5000) -layout(matrix(1:4, ncol=2)) -par(mar=c(5, 5, 5, 1)) - -# Plot all motif counts along the linear correlation coefficient -plot.scatter(region1Counts+0.5, region2Counts+0.5, log="xy", xlab=xlab, ylab=ylab, - cex.lab=2.2, cex.axis=1.8, xlim=lim, ylim=lim*rValue) -abline(0, rValue, untf=T) -abline(0, rValue*2, untf=T, lty=2) -abline(0, rValue/2, untf=T, lty=2) - -# Plot enriched and depleted motifs in red, housed in second plot -plot.scatter(region1Counts+0.5, region2Counts+0.5, log="xy", xlab=xlab, ylab=ylab, - cex.lab=2.2, cex.axis=1.8, xlim=lim, ylim=lim*rValue) -points(region1Counts[indices]+0.5, region2Counts[indices]+0.5, col="red") -abline(0, rValue, untf=T) -abline(0, rValue*2, untf=T, lty=2) -abline(0, rValue/2, untf=T, lty=2) - -# Apply and plot gc correction and loess curve -cat("Applying gc correction, rerunning poisson test...\n") -ind = which(region1Counts>5) -gc = gc[names(region2Counts)] # Reorder the indices of pwms to match input data -lo = plot.scatter(gc,log2(region2Counts/region1Counts),draw.loess=T, - xlab="gc content of motif",ylab=paste("log2(",ylab,"/",xlab,")"), - cex.lab=2.2,cex.axis=1.8,ind=ind) # This function is in plotting.r -gcCorrection = 2^approx(lo$loess,xout=gc,rule=2)$y - -# Recalculate p and q values, and filter for significant entries -pValueGC = sapply(seq(along=region1Counts),function(i) { - poisson.test(c(region1Counts[i],region2Counts[i]),r=1/gcCorrection[i])$p.value -}) -qValueGC=p.adjust(pValueGC,"fdr") -indicesGC = which(qValueGC<0.1 & abs(log2(region1Counts/region2Counts*gcCorrection))>log2(1.5)) - -# Plot gc corrected motif counts -plot.scatter(region1Counts+0.5, (region2Counts+0.5)/gcCorrection, log="xy", - xlab=xlab, ylab=paste(ylab,"(normalized)"), cex.lab=2.2, cex.axis=1.8, - xlim=lim, ylim=lim) -points(region1Counts[indicesGC]+0.5, - (region2Counts[indicesGC]+0.5)/gcCorrection[indicesGC], col="red") -abline(0,1) -abline(0,1*2,untf=T,lty=2) -abline(0,1/2,untf=T,lty=2) - -# Trim results, compile statistics and output to file -# Only does so if significant results are computed -if(length(indicesGC) > 0) { - # Calculate expected counts and enrichment ratios - cat("Calculating statistics...\n") - nullExpect = region1Counts * gcCorrection - enrichment = region2Counts / nullExpect - - # Reorder selected indices in ascending pvalue - cat("Reordering by ascending pvalue...\n") - indicesReorder = indicesGC[order(pValueGC[indicesGC])] - - # Combine data into one data frame and output to two files - cat("Splitting and outputting data...\n") - outDF = data.frame(motif=names(pValueGC), p=as.numeric(pValueGC), q=qValueGC, - stringsAsFactors=F, region_1_count=region1Counts, - null_expectation=round(nullExpect,2), region_2_count=region2Counts, - enrichment=enrichment)[indicesReorder,] - names(outDF)[which(names(outDF)=="region_1_count")]=xlab - names(outDF)[which(names(outDF)=="region_2_count")]=ylab - indicesEnrich = which(outDF$enrichment>1) - indicesDeplete = which(outDF$enrichment<1) - outDF$enrichment = ifelse(outDF$enrichment>1, - round(outDF$enrichment,3), - paste("1/",round(1/outDF$enrichment,3))) - write.table(outDF[indicesEnrich,], file=enrichTab, quote=FALSE, - sep="\t", append=FALSE, row.names=FALSE, col.names=TRUE) - write.table(outDF[indicesDeplete,], file=depleteTab, quote=FALSE, - sep="\t", append=FALSE, row.names=FALSE, col.names=TRUE) -} - -# Catch display messages and output timing information -catchMessage = dev.off() -cat("Done. Job started at:", format(startTime, "%a %b %d %X %Y."), - "Job ended at:", format(Sys.time(), "%a %b %d %X %Y."), "\n")
--- a/upload/region_motif_compare.xml Wed Nov 12 15:19:37 2014 -0500 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,32 +0,0 @@ -<tool id="region_motif_compare" name="Region Motif Count Compare"> - <description>for comparing the motif counts in different region sets</description> - <requirements> - <requirement type="set_environment">RMOTIF_PATH</requirement> - </requirements> - <command interpreter="Rscript"> - region_motif_compare.r --args \$RMOTIF_PATH $db_type $in_tab_1 $in_tab_2 $out_enriched $out_depleted $out_plots - </command> - <inputs> - <param name="in_tab_1" type="data" format="tabular" label="Region Set 1 Motif Count File"/> - <param name="in_tab_2" type="data" format="tabular" label="Region Set 2 Motif Count File"/> - <param name="db_type" type="select" label="Select Motif Database" > - <option value="t">Test Pouya Subset (hg19)</option> - <option value="p">Pouya Encode Motifs (hg19)</option> - <option value="j">Jaspar and Jolma Motifs (hg19)</option> - <option value="m">Mouse Motifs (mm9)</option> - <option value="c">Pouya, Jaspar, and Jolma Combined (hg19)</option> - </param> - </inputs> - <outputs> - <data name="out_enriched" format="tabular" label="Enriched Motifs"/> - <data name="out_depleted" format="tabular" label="Depleted Motifs"/> - <data name="out_plots" format="png" label="Motif Count Comparison Plots"/> - </outputs> - - <help> - This tools reads in two counts file and determines enriched and depleted - motifs in two different region sets based on poisson calculation with - gc correction. - </help> - -</tool> \ No newline at end of file
--- a/upload/region_motif_intersect.r Wed Nov 12 15:19:37 2014 -0500 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,102 +0,0 @@ -# Name: region_motif_intersect.r -# Description: Takes a bed file of target regions and counts intersections -# of each motif (built in rdata database) and target regions. -# Author: Jeremy liu -# Email: jeremy.liu@yale.edu -# Date: 14/07/02 -# Note: This script is meant to be invoked with the following command -# R --slave --vanilla -f ./region_motif_intersect.r --args <workingdir> <db> <inbed> <outtab> -# <workingdir> is working directory of galaxy installation -# <db> types: "t" test, "p" pouya, "j" jaspar jolma, "m" mouse -# Dependencies: none - -# Auxiliary function to concatenate multiple strings -concat <- function(...) { - input_list <- list(...) - return(paste(input_list, sep="", collapse="")) -} - -# Set common and data directories -args <- commandArgs() -workingDir = args[7] -dbDir = concat(workingDir, "/region_motif_db") -dbCode = args[8] -if (dbCode == "t") { - motifDB = concat(dbDir, "/pouya_test_motifs.bed.bgz") -} else if (dbCode == "p") { - motifDB = concat(dbDir, "/pouya_motifs.bed.bgz") -} else if (dbCode == "j") { - motifDB = concat(dbDir, "/jaspar_jolma_motifs.bed.bgz") -} else if (dbCode == "m") { - motifDB = concat(dbDir, "/mm9_motifs.bed.bgz") -} else { - motifDB = concat(dbDir, "/pouya_motifs.bed.bgz") -} - -# Set input and reference files, comment to toggle commmand line arguments -inBed = args[9] -outTab = args[10] - -# Auxiliary function to read in BED file -read_bed <- function(file) { - return(read.table(file, sep="\t", stringsAsFactors=FALSE)) -} - -startTime = Sys.time() -cat("Running ... Started at:", format(startTime, "%a %b %d %X %Y"), "...\n") - -# Load dependencies -cat("Loading dependencies...\n") -suppressPackageStartupMessages(library(Rsamtools, quietly=TRUE)) - -# Initializing hash table (as env) with motif names and loading tabix file -cat("Loading motif database and initializing hash table...\n") -motifTable = new.env() -motifTbx <- TabixFile(motifDB) - -# Loading input bed file, convert integer columns to numeric, name columns -cat("Loading region file...\n") -regionsDF = read_bed(inBed) -dfTemp = sapply(regionsDF, is.integer) -regionsDF[dfTemp] = lapply(regionsDF[dfTemp], as.numeric) -names(regionsDF)[names(regionsDF) == "V1"] = "chr" -names(regionsDF)[names(regionsDF) == "V2"] = "start" -names(regionsDF)[names(regionsDF) == "V3"] = "end" - -# Filtering regions to exclude chromosomes not in motif database -cat("Determining intersection counts...\n") -motifTbxChrs = seqnamesTabix(motifTbx) -regionsDFFilter = subset(regionsDF, chr %in% motifTbxChrs) - -# Loading regions into GRanges object and scanning motif tabix database -# Region end is incremented by 1 since scanTabix querying is inclusive for -# position start but exclusive for position end. -param = GRanges(regionsDFFilter$chr, IRanges(regionsDFFilter$start, - end=regionsDFFilter$end + 1)) -regionsIntersects = scanTabix(motifTbx, param=param) - -# Parsing result list and updating motif count hash table -cat("Parsing result list...\n") -for(regionIntersects in regionsIntersects) { - for(regionIntersect in strsplit(regionIntersects, " ")) { - intersectMotif = strsplit(regionIntersect, "\t")[[1]][4] - if(is.null(motifTable[[intersectMotif]])) { - motifTable[[intersectMotif]] = 1 - } else { - motifTable[[intersectMotif]] = motifTable[[intersectMotif]] + 1 - } - } -} - -# Converting motif count hash table to an integer vector for output -counts = integer(length = length(ls(motifTable))) -names(counts) = ls(motifTable) -for(motifName in ls(motifTable)) { - counts[motifName] = as.integer(motifTable[[motifName]]) -} - -# Outputting intersection counts to tab delineated file -cat("Outputting to file...\n") -write.table(counts, outTab, quote=FALSE, sep="\t", row.names=TRUE, col.names=FALSE) -cat("Done. Job started at:", format(startTime, "%a %b %d %X %Y."), - "Job ended at:", format(Sys.time(), "%a %b %d %X %Y."), "\n")
--- a/upload/region_motif_intersect.xml Wed Nov 12 15:19:37 2014 -0500 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,27 +0,0 @@ -<tool id="region_motif_intersect" name="Region Motif Intersect"> - <description>for computing the motifs that lie inside a region set</description> - <requirements> - <requirement type="set_environment">RMOTIF_PATH</requirement> - </requirements> - <command interpreter="Rscript"> - region_motif_intersect.r --args \$RMOTIF_PATH $db_type $in_bed $out_tab - </command> - <inputs> - <param name="in_bed" type="data" format="bed" label="Input BED File" /> - <param name="db_type" type="select" label="Select Motif Database" > - <option value="t">Test Pouya Subset (hg19)</option> - <option value="p">Pouya Encode Motifs (hg19)</option> - <option value="j">Jaspar and Jolma Motifs (hg19)</option> - <option value="m">Mouse Motifs (mm9)</option> - </param> - </inputs> - <outputs> - <data name="out_tab" format="tabular" /> - </outputs> - - <help> - This tool computes the motifs and the number of motifs that intersect - any region in a input set of regions. - </help> - -</tool> \ No newline at end of file
--- a/upload/tool_dependencies.xml Wed Nov 12 15:19:37 2014 -0500 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,6 +0,0 @@ -<?xml version="1.0"?> -<tool_dependency> - <set_environment version="1.0"> - <environment_variable name="RMOTIF_PATH" action="set_to">$REPOSITORY_INSTALL_DIR</environment_variable> - </set_environment> -</tool_dependency> \ No newline at end of file