fasta file extraction

Question

[amitpande74]

Hi, I have a fasta file formatted like this, product of paired end:

test_MAPQ.fasta->chr10:141146-141296
test_MAPQ.fasta:atgctcccattccaaatgagagtaattggctaaaacaaaggggctacaggtcccatacaagtccaaaacccaacagggcagtcattaaatcttTTCTAAttttaatttttattttatttgaagttctggggtacatgttcaggatgtata
test_MAPQ.fasta->chr10:142926-143076
--
test_MAPQ.fasta->chr10:146793-146943
test_MAPQ.fasta:gccAAATCATTACTTTTGAAGAAATAGTTAACAATGATTATTTCTTTTTGAATGACaataaattttattaataagttaaacatatttatatgtaatgtaaattttttGTATcgggtgcagtggttcatgcccgttatcctagcactttgg
test_MAPQ.fasta->chr10:146870-147020
test_MAPQ.fasta:aaacatatttatatgtaatgtaaattttttGTATcgggtgcagtggttcatgcccgttatcctagcactttgggaggccaaggtgttaatattgcttgagcaggggagtttgagaccagcctgggaaacatggtgaaacctcatatctac

I want an output where only both the pairs are included in the final results.

test_MAPQ.fasta->chr10:146793-146943
test_MAPQ.fasta:gccAAATCATTACTTTTGAAGAAATAGTTAACAATGATTATTTCTTTTTGAATGACaataaattttattaataagttaaacatatttatatgtaatgtaaattttttGTATcgggtgcagtggttcatgcccgttatcctagcactttgg
test_MAPQ.fasta->chr10:146870-147020
test_MAPQ.fasta:aaacatatttatatgtaatgtaaattttttGTATcgggtgcagtggttcatgcccgttatcctagcactttgggaggccaaggtgttaatattgcttgagcaggggagtttgagaccagcctgggaaacatggtgaaacctcatatctac

and the rest are ignored (like the upper pairs ending in dashes). Is there a tool which can filter out the results ? Kindly help.

[cpad0112]

what is the basis of pairing from OP example? Number of lines before -- ? amitpande74.

$ awk -v RS="--" -v ORS="--" 'NF>3 {print}' test.fa

test_MAPQ.fasta->chr10:146793-146943
test_MAPQ.fasta:gccAAATCATTACTTTTGAAGAAATAGTTAACAATGATTATTTCTTTTTGAATGACaataaattttattaataagttaaacatatttatatgtaatgtaaattttttGTATcgggtgcagtggttcatgcccgttatcctagcactttgg
test_MAPQ.fasta->chr10:146870-147020
test_MAPQ.fasta:aaacatatttatatgtaatgtaaattttttGTATcgggtgcagtggttcatgcccgttatcctagcactttgggaggccaaggtgttaatattgcttgagcaggggagtttgagaccagcctgggaaacatggtgaaacctcatatctac

You can also do awk 'BEGIN{RS="--"} NF>3 {print} test.fa but it would create empty lines before and after the sequences.