RSEM error (with Trinity script) when using filtered assembly

Question

[swaegers.janne]

Hi everyone,

I'm running RSEM on a sample using a de novo assembly where part of the contigs were removed because it were duplicates (I used CD-HIT_EST clustering method)...Now it appears that this does not work. It does work when I use the original assembly fasta file. Is there a reason for this?

These is part of the error I get when using the assembly where duplicates were removed:

RSEM can not recognize reference sequence name 183617!
Error, cmd: rsem-calculate-expression  --paired-end   -p 4    --no-bam-output --bam FCH7F53BBXX-HKISCsixEAAARAAPEI-208_L7_one.bowtie.bam /ddn1/vol1/site_scratch/leuven/315/vsc31552/Trinity95.fasta.RSEM FCH7F53BBXX-HKISCsixEAAARAAPEI-208_L7_one  died with ret: 65280 at /data/leuven/315/vsc31552/Trinity/trinityrnaseq-2.2.0/util/align_and_estimate_abundance.pl line 743.

My script is the following:

/data/leuven/315/vsc31552/Trinity/trinityrnaseq-2.2.0/util/align_and_estimate_abundance.pl --seqType fq  \
      --left /scratch/leuven/315/vsc31552/RNAseq/F16FTSEUHT0283ISCsixE/cleanreads/P-02/FCH7F53BBXX-HKISCsixEAABRAAPEI-209_L7_forward_paired.fq.gz --right /scratch/leuven/315/vsc31552/RNAseq/F16FTSEUHT0283ISCsixE/cleanreads/P-02/FCH7F53BBXX-HKISCsixEAABRAAPEI-209_L7_reverse_paired.fq.gz \
      --transcripts /scratch/leuven/315/vsc31552/RNAseq/Trinity95.fasta  \
      --est_method RSEM  --aln_method bowtie \
      --trinity_mode --prep_reference --output_dir /scratch/leuven/315/vsc31552/RNAseq/F16FTSEUHT0283ISCsixE/cleanreads/P-02/
echo "================="
exit 0

Thanks in advance!

Janne

[st.ph.n]

is 183617 in your Trinity95.fasta? the wrapper script should have prepped/indexed your reference sequence (given you used the parameter), but 183617 is not a Trinity fasta header, and you are specifying 'trinity mode'. Did you fasta headers change?

Answer 1

[swaegers.janne]

Hi! Thanks for looking into this!

Both my filtered fasta file (with which the analysis does not work) as well as the original (with which I don't get an error) have a transcript which incorporates that number ("TRINITY_DN183617_c0_g1_i1").

The fasta headers haven't changed: head of filtered fasta:

TRINITY_DN88414_c0_g1_i1 len=221 path=[247:0-76 248:77-79 249:80-100 250:101-103 251:104-220] [-1, 247, 248, 249, 250, 251, -2] CTAAGGTAAGGTTTTATAGCTGAGGACATGTTGCAGTTAGAGCCGTAGCCGATTTTCGTT TGTAGGAATTATTTTGGCGGTAATAAGTAGTGGGTTGCGTGTTTAACGCTACTCTTTCGG AGTTGGAAGGTTCTGGCTGAGGTTGGTATGTACGAATGATTGGCAAAATAACCTATTTGG AGATTATTGCTAAATGCTCATATTACTCCCACTATAATCTC TRINITY_DN88414_c1_g2_i1 len=322 path=[446:0-108 447:109-132 470:133-159 467:160-198 471:199-203 468:204-221 465:222-245 464:246-321] [-1, 446, 447, 470, 467, 471, 468, 465, 464, -2] AAACAACCTATTTGGAGATTATTACTAAATGGGGTTATCATATTACTCCCACTATAATCT CCAGATGTATCAAAAGCTAGAAGTAAGTTAGTCGTATATTTCTCTCAAAACTATGTTAAC GCCACTGATGTAATGGAGAGAAGTTATTGATCAAAGAGATATTGAAACTAATTATGGCTG GAGGAGCAATACTATTACGGGAATACCTTAAGGTAATTCTGTGCTGACGAAAATGGGAAG

head of original fasta:

TRINITY_DN88414_c0_g1_i1 len=221 path=[247:0-76 248:77-79 249:80-100 250:101-103 251:104-220] [-1, 247, 248, 249, 250, 251, -2] CTAAGGTAAGGTTTTATAGCTGAGGACATGTTGCAGTTAGAGCCGTAGCCGATTTTCGTT TGTAGGAATTATTTTGGCGGTAATAAGTAGTGGGTTGCGTGTTTAACGCTACTCTTTCGG AGTTGGAAGGTTCTGGCTGAGGTTGGTATGTACGAATGATTGGCAAAATAACCTATTTGG AGATTATTGCTAAATGCTCATATTACTCCCACTATAATCTC TRINITY_DN88414_c0_g2_i1 len=221 path=[252:0-76 253:77-100 254:101-103 255:104-220] [-1, 252, 253, 254, 255, -2] CTAAGGTAAGGTTTTATAGCTGAGGACATGTTGCAGTTAGAGCCGTAGCCGATTTTCGTT TGTAGGAATTATTTTGGTGGTAATAAGTAGTGGGTTGCGTGTTTAACGCTACTCTTTCGG AGTTGGAAGGTTCTGGCTGAGGTTGGTATGTACGAATGATTGGCAAAATAACCTATTTGG AGATTATTGCTAAATGCTCATATTACTCCCACTATAATCTC

[swaegers.janne]

Would anyone know what is causing this error? Thanks a lot!