Match patterns between two files and extract certain range of strings

Hi,

I need help to match patterns from between two different files and extract region of strings.

inputfile1.fa

>l-WR24-1:1
GCCGGCGTCGCGGTTGCTCGCGCTCTGGGCGCTGGCGGCTGTGGCTCTACCCGGCTCCGG
GGCGGAGGGCGACGGCGGGTGGTGAGCGGCCCGGGAGGGGCCGGGCGGTGGGGTCACGTG
CGGCGGGCGGGGCGCGGGCTGACCCAGCTGTGCCCGCAGGCGCCCGGGCGGGCCGGGGGC
CGTGGCGGAGGAGGAGCGCTGCACGGTGGAGCGTCGGGCCGACCTCACCTACGCGGAGTT
CTACCACAAAGTGGACTTGCCCTTCCAGGAGTATGTGGAGCAGCTGCTGCACCCCCAGGA
CCCCACCTCCCTGGGCAATGACACCCTGTACTTCTTCGGGGACAACAACTTCACCGAGTG
GGCCTCTCTCTTTCGGCACTACTCCCCACCCCCATTTGGCCTGCTGGGAACCGCTCCAG
>l-ZF385A-2:1
CAGAATGTGGGTGAGGGTGGCGCCTATGAGGCTGAGCTTCGGGTCACCGCCCCTCCAGAG
GCTGAGTACTCAGGACTCGTCAGACACCCAGGGGTGAGATGAGACTCTCGAGTGGGATTT
GGGAGGATACCCCTCTAGAGGGGACACCAAAACCTGACCAGTGCCCACCCCATCTCCAGA
ACTTCTCCAGCCTGAGCTGTGACTACTTTGCCGTGAACCAGAGCCGCCTGCTGGTGTGTG
ACCTGGGCAACCCCATGAAGGCAGGAGCCAGTCTGTGGGGTGGCCTTCGGTTTACAGTCC
CTCATCTCCGGGACACTAAGAAAACCATCCAGTTTGACTTCCAGATCCTCAGGTAGGGAG
TGAGTGTGTCTAGGCTGGGGCTGAGCTGGGGACGGAAGGGAGGGCTGGGCGCCATTCTCA
CTGGCTGCACTCCAGCACCTCAGTCTTGCCTCCATCCCACAGCAAGAATCTCAACAACTC
GCAAAGCGACGTGGTTTCCTTTCGGCTCTCCGTGGAGGCTCAGGCCCAGGTCACCCTGAA
CGGGTCAGTGCCAGGCAAAATGGGGTCT
>l-YJC-1:1
GTCCCGCCCTCGCATGCGCCTGGTGGTCACCGCGGACGACTTTGGTTACTGCCCGCGACG
CGATGAGGGTATCGTGGAGGCCTTTCTGGCCGGGGCTGTGACCAGCGTGTCCCTGCTGGT
CAACGGTGCGGCCACGGAGAGCGCGGCGGAGCTGGCCCGCAGGCACAGCATCCCCACGGG
CCTCCACGCCAACCTGTCCGAGGGCCGCCCCGTGGGTCCGGCCCGCCGTGGCGCCTCATC
GCTGCTCGGCCCGG

inputfile2.txt

l-WR24-1:1	1	71
l-WR73-7:4	28	506
l-WR86-1:1	140	138
l-YJC-1:1	1	161
l-YJC-1:1	1	165
l-ZFP57-11:1	1	991
l-ZF320-1:5	5	6031
l-ZMYND10-2:1	5	253
l-ZF329-4:1	151	5704
l-ZF708-1:1	195	3744
l-ZF843-3:1	14	1053
l-ZF385A-2:1	33	105
l-ZF843-3:2	4	235

The output file as below. It should contain the strings extracted from inputfile1.fa according to start and stop numbers (in blue) indicated in inputfile2.txt as shown above.

>l-WR24-1:1	1	71
GCCGGCGTCGCGGTTGCTCGCGCTCTGGGCGCTGGCGGCTGTGGCTCTACCCGGCTCCGG
GGCGGAGGGCG
>l-ZF385A-2:1	33	105
TGAGCTTCGGGTCACCGCCCCTCCAGAGGCTGAGTACTCAGGACTCGTCAGACACCCAGG
GGTGAGATGAGAC
>l-YJC-1:1	1	161
GTCCCGCCCTCGCATGCGCCTGGTGGTCACCGCGGACGACTTTGGTTACTGCCCGCGACG
CGATGAGGGTATCGTGGAGGCCTTTCTGGCCGGGGCTGTGACCAGCGTGTCCCTGCTGGT
CAACGGTGCGGCCACGGAGAGCGCGGCGGAGCTGGCCCGCA
>l-YJC-1:1	1	165
GTCCCGCCCTCGCATGCGCCTGGTGGTCACCGCGGACGACTTTGGTTACTGCCCGCGACG
CGATGAGGGTATCGTGGAGGCCTTTCTGGCCGGGGCTGTGACCAGCGTGTCCCTGCTGGT
CAACGGTGCGGCCACGGAGAGCGCGGCGGAGCTGGCCCGCAGGCA

I have tried to match these two files using bioinformatics tools like seqtk and bedtools but they did not give me what I want and some of their data requirements just don't fit mine. I am still learning awk and I want to do this in awk. Below are one of the codes that I did to match the two files but failed and I do not know how to extract certain regions of characters.

awk 'FNR == NR{A[$1]; next}/^\>/{f = (substr($1,2) in A) ? 1 : 0}f' inputfile2.txt inputfile1.fa

Would appreciate your kind help. thanks

a bit verbose, but a possible starter.
awk -f bunny.awk inputfile2.txt inputfile1.fa where bunny.awk is:

function printRec() {
   #print a[f], s[f], e[f]
   for ( i in s) {
      split(i,t, OFS)
      if (f == ">" t[1])
        print ">" i ORS substr(a[f],s,e-s+1)
   }
   f=""
   split("",a)
}
FNR==NR {
   idx=$1 OFS $2 OFS $3
   s[idx]=$2
   e[idx]=$3
   next
}
/>/ && f {
   printRec()
}
f { a[f]=(f in a)?a[f] $1:$1 }
/^>/ { f=$1 }
END { printRec() }

results in:

>l-WR24-1:1 1 71
GCCGGCGTCGCGGTTGCTCGCGCTCTGGGCGCTGGCGGCTGTGGCTCTACCCGGCTCCGGGGCGGAGGGCG
>l-ZF385A-2:1 33 105
TGAGCTTCGGGTCACCGCCCCTCCAGAGGCTGAGTACTCAGGACTCGTCAGACACCCAGGGGTGAGATGAGAC
>l-YJC-1:1 1 161
GTCCCGCCCTCGCATGCGCCTGGTGGTCACCGCGGACGACTTTGGTTACTGCCCGCGACGCGATGAGGGTATCGTGGAGGCCTTTCTGGCCGGGGCTGTGACCAGCGTGTCCCTGCTGGTCAACGGTGCGGCCACGGAGAGCGCGGCGGAGCTGGCCCGCA
>l-YJC-1:1 1 165
GTCCCGCCCTCGCATGCGCCTGGTGGTCACCGCGGACGACTTTGGTTACTGCCCGCGACGCGATGAGGGTATCGTGGAGGCCTTTCTGGCCGGGGCTGTGACCAGCGTGTCCCTGCTGGTCAACGGTGCGGCCACGGAGAGCGCGGCGGAGCTGGCCCGCAGGCA

Try also

awk '
NR==FNR         {PAT[$1,$2,$3]
                 next
                }
                {IX  = $1
                 L1  = length ($1) + 1
                 $1 = $1 "|"
                 $0 = $0
                 for (p in PAT) {split (p, T)
                                 if (IX == T[1]) print RS p ORS substr ($0, T[2]+L1, T[3]-T[2]+1)
                                }
                }
' SUBSEP="\t" inputfile2.txt   RS=">"  OFS="" inputfile1.fa

>l-WR24-1:1    1    71
GCCGGCGTCGCGGTTGCTCGCGCTCTGGGCGCTGGCGGCTGTGGCTCTACCCGGCTCCGGGGCGGAGGGCG
>l-ZF385A-2:1    33    105
TGAGCTTCGGGTCACCGCCCCTCCAGAGGCTGAGTACTCAGGACTCGTCAGACACCCAGGGGTGAGATGAGAC
>l-YJC-1:1    1    161
GTCCCGCCCTCGCATGCGCCTGGTGGTCACCGCGGACGACTTTGGTTACTGCCCGCGACGCGATGAGGGTATCGTGGAGGCCTTTCTGGCCGGGGCTGTGACCAGCGTGTCCCTGCTGGTCAACGGTGCGGCCACGGAGAGCGCGGCGGAGCTGGCCCGCA
 >l-YJC-1:1    1    165
GTCCCGCCCTCGCATGCGCCTGGTGGTCACCGCGGACGACTTTGGTTACTGCCCGCGACGCGATGAGGGTATCGTGGAGGCCTTTCTGGCCGGGGCTGTGACCAGCGTGTCCCTGCTGGTCAACGGTGCGGCCACGGAGAGCGCGGCGGAGCTGGCCCGCAGGCA

If you really really need the output lines 60 chars in length, use

                 if (IX == T[1])    {print RS p
                                     TMP = substr ($0, T[2]+L1, T[3]-T[2]+1)
                                     PTR = 1
                                     while (PTR < length (TMP))    {print substr (TMP, PTR, 60)
                                                                    PTR += 60
                                                                   }
                                    }

if you need to "fold" the lines at 60 char width, pipe the output to fold :

awk .... | fold -w 60

Hi vgersh99,

Your codes work like charm. Thanks a million.

--- Post updated at 03:59 AM ---

Hi RudiC,

Your codes work great on my real data too. Thanks a million

--- Post updated at 04:43 AM ---

Ok, got it... thanks