comparison dnp-smooth.sh @ 0:448204d12325 draft default tip

"planemo upload commit 1a32efb8343938e8d49190003f251c78b5a58225-dirty"
author erinija
date Fri, 01 May 2020 12:13:00 +0000
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:448204d12325
1 #!/bin/sh
2
3 if test "$#" -ne 4; then
4
5 echo ""
6 echo " CALL "
7 echo " sh dnp-smooth.sh smoothing-input.tabular winsize trim smoothed-output.tabular"
8 echo ""
9 echo " INPUT "
10 echo " smoothing-input.tabular - dinucleotide frequency profiles-patterns to smooth"
11 echo " winsize - size of averaging window, suggested optimal value =3"
12 echo " trim - how many noisy points to remove from both ends of the profile suggested =4"
13 echo ""
14 echo " OUTPUT "
15 echo " smoothed-output.tabular - original series smoothed "
16 echo ""
17 echo " DESCRIPRION"
18 echo " Applies smoothing on dinucleotide profiles. Smoothing reduces noise and enhances"
19 echo " a representation of the dinucleotide frequency profiles. Smoothing is performed by"
20 echo " moving average with a chosen window size (optimal winsize=3). Smoothing script is based on"
21 echo " a shell wrapper of the call to dnp-fourier tool which computes a periodogram"
22 echo " of a series given as a numerical column and has parameters:"
23 echo " dnp-fourier -f input -o output -n {normalization 0|1|2} -l length_of_smoothing_window -t {type_of_output 1|2|3}"
24 echo " The parameters control a type of normalization and output as follows:"
25 echo " Normalization"
26 echo " 0 base normalization subtracts mean"
27 echo " 1 linear normalization removes linear trand "
28 echo " 2 quadratic normalization removes quadratic trend"
29 echo " Output type"
30 echo " 1 normalization outputs normalized original series"
31 echo " 2 smoothing outputs smoothed original series"
32 echo " 3 Fourier transform outputs periofogram"
33 echo ""
34 echo " Example of input table"
35 echo " pos AA AC AG AT CA CC CG CT GA GC GG GT TA TC TG TT WW SS RR YY"
36 echo " -73 0.05664 0.0657 0.06966 0.03644 0.08026 0.09484 0.0362 0.09086 0.07084 0.04032 0.07318 0.06466 0.03862 0.06838 0.05722 0.05602 0.18772 0.24454 0.27032 0.3101"
37 echo " -72 0.0668 0.06476 0.0753 0.04282 0.07022 0.08034 0.03534 0.081 0.07222 0.03512 0.06774 0.0598 0.03934 0.07496 0.06628 0.06784 0.2168 0.21854 0.28206 0.30414"
38 echo " -71 0.063 0.0621 0.07668 0.04316 0.06926 0.07264 0.03316 0.07992 0.07546 0.03498 0.07306 0.06406 0.04182 0.07374 0.06874 0.06812 0.2161 0.21384 0.2882 0.29442"
39 echo " -70 0.0624 0.0643 0.07214 0.04424 0.0642 0.06998 0.03472 0.07718 0.0723 0.03982 0.07472 0.06818 0.04282 0.07674 0.06864 0.06754 0.217 0.21924 0.28156 0.29144"
40 echo " -69 0.0622 0.06456 0.074 0.0426 0.0661 0.07114 0.03414 0.08016 0.0703 0.03786 0.07118 0.06754 0.0421 0.07712 0.06988 0.06904 0.21594 0.21432 0.27768 0.29746"
41 echo " ..."
42 echo ""
43 echo " Output table is an original input table but smoothed by moving average with given window size."
44 echo ""
45 echo " REQUIREMENT"
46 echo " dnp-fourier installed"
47 echo " conda install -c bioconda dnp-fourier"
48 echo ""
49
50
51 exit 1
52 fi
53
54 # input file name
55 name=$1
56 swindow=$2
57 trim=$3
58 out=$4
59
60 call=dnp-fourier
61
62 ## get the nucleotides from the header
63 dinucleotides=`head -n1 ${name} | sed 's/pos//'`
64
65
66 ## save the position information
67 posi=`awk '{ for (i=1; i<=NF; i++) if($i=="pos") print i; exit}' ${name}`
68 awk -v k=${posi} '{print $k}' $name | grep -v pos > posfile
69
70
71 for di in ${dinucleotides}
72 do
73 # column number of dinucleotide
74 i1=`awk -v name=$di '{ for (i=1; i<=NF; i++) if($i==name) print i; exit}' ${name}`
75 awk -v k=${i1} '{print $k}' $name | grep -v ${di} > temp.${di}
76
77 # add centered sequence position
78 paste posfile temp.${di} | tr "\t" " " > smoothprep
79
80 # perform smoothing
81 ${call} -f smoothprep -t 2 -l ${swindow} -o cps.${di}
82
83
84 # echo ${di} > cp.${di}
85 cat cps.${di} | tail -n +$(($trim+1)) | head -n -$(($trim+1)) > temp.${di}
86
87 echo "pos" > positions
88 cat temp.${di} | awk '{print $1}' >> positions
89 echo ${di} > cp.${di}
90 cat temp.${di} | awk '{print $2}' >> cp.${di}
91
92 done
93
94 #echo "pos" ${dinucleotides} | tr " " "\t" > ${out}
95 paste positions cp.* > ${out}
96
97 rm temp.* cp.* cps.* posfile smoothprep positions