convert-sam-for-rsem -p 12 --memory-per-thread 20G bowtie2.bam bowtie2.bam.for_rsem

[tomas-pluskal]

Hi,

I am trying to use RSEM to estimate the expression of transcripts assembled by Trinity. I am using the Trinity's align_and_estimate_abundance.pl script. The script fails at the convert-sam-for-rsem step with an error, which I can reproduce by running this command manually:

convert-sam-for-rsem -p 12 --memory-per-thread 20G bowtie2.bam bowtie2.bam.for_rsem

This command produced the following error:

samtools sort -n -@ 12 -m 20G -o bowtie2.bam.for_rsem.tmp.bam bowtie2.bam
[bam_sort_core] merging from 0 files and 12 in-memory blocks...

rsem-scan-for-paired-end-reads 12 bowtie2.bam.for_rsem.tmp.bam bowtie2.bam.for_rsem.bam
.Number of first and second mates in read HWI-1KL163:66:C29PVACXX:8:1101:1028:68763_forward's full alignments (both mates are aligned) are not matched!

"rsem-scan-for-paired-end-reads 12 bowtie2.bam.for_rsem.tmp.bam bowtie2.bam.for_rsem.bam" failed! Plase check if you provide correct parameters/options for the pipeline!

What is wrong? The raw paired reads were downloaded from SRA and look fine. Below is the output of samtools flagstat bowtie2.bam:

481808982 + 0 in total (QC-passed reads + QC-failed reads)
325773752 + 0 secondary
0 + 0 supplementary
0 + 0 duplicates
481808982 + 0 mapped (100.00% : N/A)
156035230 + 0 paired in sequencing
78017615 + 0 read1
78017615 + 0 read2
156035230 + 0 properly paired (100.00% : N/A)
156035230 + 0 with itself and mate mapped
0 + 0 singletons (0.00% : N/A)
0 + 0 with mate mapped to a different chr
0 + 0 with mate mapped to a different chr (mapQ>=5)

Tomas

[Lcornet]

Dear Tomas,

Any update on this? I got the same error with a apparently nothing wrong with the reads. Did you manage to solve this ?

[tomas-pluskal]

@Lcornet Honestly, I don't remember, the issue is from 2018.. We now use kallisto for expression analysis, it is much faster.