bbmap inflating read count and not finding one sequence after header

[mforthman]

I'm using bbmap to map transcriptome reads to a set of target loci. I'm working with 12 samples with pair-end reads and 1 sample with single-end reads, all from NCBI's SRA. I'm having no problems with bbmap reading paired-end data and completing analyses correctly. It's the one sample with single-end reads that's causing two issues:

1. The first issue is that the input fasta file only has 288915 reads. I have confirmed this with grep ">" file.fasta | wc -l. However, bbmap reports "Reads used: 308655". I have no idea why the read count is inflated; again, this is not an issue with the paired-end data.

2. bbmap fails to recognize a sequence immediately after the fasta header: "Warning: A fasta header with no sequence was encountered: SRR768524.9631" The sequence in question is formatted exactly like all others in the file and I have checked the EOL, which is fine. Below is what a snippet of the fasta file looks like, with the 3rd sequence being the problematic one for bbmap.

>SRR768524.9629
CTATCAAAGGGAAATCCCGCTGGCCTGCTATCATACAGTCTTGAACCTCCACATGCAATATCAGCTGAATCTCCAACGTGGCCGCTGACAAATGGAGTAACTACGACTGCCAAAACGAAAGCGCGACCTCCTTTCCATCCCATGGGTAATTGGAGTCTTTGAGGAAATCCACATCGGCCAGCCTCTGAATAATGGAATGGTTCCTTCTGGTTGAGTGCACTGTTTATTGAAGTGTAAAGAGACCTGAATCCTTCTTGGTCATGGATAAAGAGGGGAGAATCTGTTGAACTTCTTTCCCACACATTAACACCCTCAGTCAATTTTACTGGGAAACGGTCAATTTCAAATAAGCTAGTTTTGCTTGGTCATAAGTGAGGAAATGTTCATCAGAATCATATTTCGGTCCAATGAATACTCTCACAATGGCATCATCAGCTTTT
>SRR768524.9630
ATAATGCAATTATAGATTGTTGGAGTGCAGGTAAAGCTACCACTGTTATGATTAAAGATAATCCAAAGGTTGAAATTCTTGATGTAGAAGATGTTAAGGTTGGAAAGATAAGACAATTTTGTGAGTTGGACTTGGCATTGAACATGGCCTTACGAAAGTATTTTGGTAGTGTGTTTGATAAAATGGCAGTTACATCTAATGAAACGCCGTGGAAAGTTGCTTGGAATCCATATTTTATGCCTCATCACATCGTGGCGATAGAGAACGACAAGTACGATGTCTTTTGTATAGATGTGAAAAGAATGGATAAGAATTTACCAGTCCAATTCACTGAGATATTGTGT
>SRR768524.9631
TGTTACTGGGTAGGGCTGTGGCACTGGGACCTTGACTGGATAGGGTACATGCTTCTCGACTGGTACTGGGTATGGCTTGGGAATGTGGACAGGGTATGGCCTATCGACTGGGACCTTCACGGGATAGGGAACTTTCTTCTCGATATGGACTGGATAAGGTACTGGTACCTCCTTGTGGATGGTGATAGCTTTAATGTGGCCGTGTTCCTCATGTCCTCCGAGTTCATAGCCACCTCCATATCCGTATCCTCCTAAGTCATGTCCACCTCCATATCCTCCATGTCCACCGCCGTATCCTCCAAGCTCATAACCACCTCCATATCCTAAGCCAAGAAGTCCTCGTTTCTCCTGCTTCTTGTCATCTGTTGGTGCCGCTGCTTTGTCGGTCTTGGATTCGGCCTTCTTCTCTTCAGCTGATGCTGTGGCAAGCAGTGCCAACAGCCCTACCCACAGTACCTTGGATTGCATTGTTGAGTCGTGGTGTGGTCGGCGTCTCCCAA
>SRR768524.9632
AATTCCCAACGACCAAGTATCTGAACATGAGTGGATCAATGCTGAGATCCTGCCTGCTACTGCTATTCCTAGCTTACTGTGTGTCCTGCTATAGAAGTCATGTTCCTAGAGGCGGGAGTTACTCTCTACCGCCTGGAGTTAATCCAACATTCCCAGGAAGGAACCAAGGACTGCCTCCGGCTTATCATGGAAAATTCAAGAGATCACTGGAAGGAGGTTTAGAACCTGAAGATGGTGGTGTCCTTGCAGTTGATGAACCTGCTGATTATCTGAAAGTCAAAAGGTCAGTGGAAGATGTTGAAGGTGAATTCCTTGTGAACGAAGAACCTCAAGAATTTGAGACACTGAGAGCGCGCCGTGACGTCAGAATAATTCATCCAACT
>SRR768524.9633
TTCCAAACTGTCGATTCATGATGTACACAATACCAAAAAAGGCAAATAAGAAATAAAAGT

I'm at a loss for figuring out how to resolve these issues. I appreciate any help in getting this to run properly on this last fasta file.

[GenoMax]

Is there a specific reason you converted these reads to fasta format? If this is data from SRA then you should be able to map the fastq reads directly.

You may be getting secondary alignments and that may be the reason why your read count seems inflated.