Incorrect codon sequences reported in kissplice2reftranscriptome output?

Question

[andres.firrincieli]

Dear Kissplice users,

after running kissplice2reftranscriptome

kissplice2reftranscriptome -b assemblychec095n10_okay_okalt_newnames.fasta.transdecoder.bed results_mDr1_EcoC_mDr2_EcoC_mDr3_EcoC_WW1_EcoC_WW2_EcoC_WW3_EcoC_k25_coherents_type_0a.fa -t out_blat_SNP_on_transcripts_80.psl -s kissDE_output_C_pvalue05

and reading the mainOutput[timestamp].tsv file, somenthing is not clear for me. As reported in the UserGuide, in the columns Codon_1 and Codon_2 should be reported
the SNP(s) from the upper and lower paths respectively. However, the codons reported in both columns (Codon_1 & Codon_2) are the same. This happen for all the SNPs; is this normal?

Thanks in advance for your time,

Best

Answer 1

[vincent.lacroix]

Dear Andres, Thanks for sending the files. It seems that the problem comes from the fact that you used a different kmer value in KisSplice and KisSplice2RefTranscriptome. We are not catching this error correctly in kissplice2reftranscriptome and it was indeed not obvious to know what the problem was. Thanks for reporting this bug. In the next release, we will clarify the manual, and add an error message.

Using the current version, the solution for you is simply to add the option -K 25 in kissplice2reftranscriptome

Best,

Vincent

[andres.firrincieli]

Dear Vincent,

thanks a lot for your help. I will try the single -K solution.

Best, Andrea

Answer 2

[vincent.lacroix]

Dear Andres,

Columns Codon 1 and Codon 2 should be distinct, otherwise there would be no SNP. Could you send me a sample of your data so that we try to reproduce your bug ? You can send directly at vincent dot lacroix at univ-lyon1 dot fr

Thanks !

Vincent