Segmentation fault with kallisto quant

[bbimber]

Hello,

I'm trying to run kallisto quant with the pseudobam option, and it is consistently running into a segfault error. Do you have suggestions on debugging that I could try? Below is the command and output:

kallisto quant -i <IDX> -o <OUTPUT> --gtf <GTF> --pseudobam -t 8 --verbose <10X_FQ1> <10X_FQ2>

and the end of the output:

[quant] done
[quant] processed 176,394,303 reads, 909,387 reads pseudoaligned
[quant] estimated average fragment length: 273.2
[   em] quantifying the abundances ... done
[   em] the Expectation-Maximization algorithm ran for 97 rounds
[  bam] writing pseudoalignments to BAM format .. /var/spool/slurmd/job39236197/slurm_script: line 118: 33371 Segmentation fault      (core dumped) $KALLISTO quant -i $RhSup_IDX -o $OUT --gtf $RhSup_GTF --pseudobam -t $THREADS --verbose $FQ1 $FQ2

thanks for any help.

[Yenaled]

What version of kallisto are you running?

One possibility I can think of is that your GTF file is malformed (i.e. the transcript IDs in your GTF file don't match the transcript IDs in your index).

[bbimber]

The version is the latest, 0.50.1.

The GTF is a bit contrived, but i think it's valid. My reference space is about 30 coding sequences. I created the GTF by making one gene feature and on transcript feature for each reference, where that feature extends one 1 to the length of the reference. I'm not sure if this is helpful, but this is the actual code, which takes the FASTA FAI index, iterates it and writes out:

	while IFS=$'\t' read -r -a myArray
	do
		SEQ_NAME="${myArray[0]}"
		SEQ_LEN="${myArray[1]}"
		echo -e $SEQ_NAME"\tnimble\tgene\t1\t"$SEQ_LEN"\t.\t+\t.\tgene_id "\"$SEQ_NAME"\"; gene_name \""$SEQ_NAME"\"; gene_biotype = \"protein_coding\";" >> $GTF
		echo -e $SEQ_NAME"\ttnimble\ttranscript\t1\t"$SEQ_LEN"\t.\t+\t.\tgene_id "\"$SEQ_NAME"\"; gene_name \""$SEQ_NAME"\"; transcript_id = \""$SEQ_NAME"\"; gene_biotype = \"protein_coding\";" >> $GTF
	done < $FAI_FILE

the result looks something like this:

CCR7_NM_001032884	nimble	gene	1	1137	.	+	.	gene_id "CCR7_NM_001032884"; gene_name "CCR7_NM_001032884"; gene_biotype = "protein_coding";
CCR7_NM_001032884	nimble	transcript	1	1137	.	+	.	gene_id "CCR7_NM_001032884"; gene_name "CCR7_NM_001032884"; transcript_id = "CCR7_NM_001032884"; gene_biotype = "protein_coding";
CD3D_XM_015115817	nimble	gene	1	2264	.	+	.	gene_id "CD3D_XM_015115817"; gene_name "CD3D_XM_015115817"; gene_biotype = "protein_coding";
CD3D_XM_015115817	nimble	transcript	1	2264	.	+	.	gene_id "CD3D_XM_015115817"; gene_name "CD3D_XM_015115817"; transcript_id = "CD3D_XM_015115817"; gene_biotype = "protein_coding";
CD3D_XM_015115818	nimble	gene	1	2249	.	+	.	gene_id "CD3D_XM_015115818"; gene_name "CD3D_XM_015115818"; gene_biotype = "protein_coding";
CD3D_XM_015115818	nimble	transcript	1	2249	.	+	.	gene_id "CD3D_XM_015115818"; gene_name "CD3D_XM_015115818"; transcript_id = "CD3D_XM_015115818"; gene_biotype = "protein_coding";
CD3E_XM_015115816	nimble	gene	1	1549	.	+	.	gene_id "CD3E_XM_015115816"; gene_name "CD3E_XM_015115816"; gene_biotype = "protein_coding";
CD3E_XM_015115816	nimble	transcript	1	1549	.	+	.	gene_id "CD3E_XM_015115816"; gene_name "CD3E_XM_015115816"; transcript_id = "CD3E_XM_015115816"; gene_biotype = "protein_coding";
CD3E_XM_028834033	nimble	gene	1	1258	.	+	.	gene_id "CD3E_XM_028834033"; gene_name "CD3E_XM_028834033"; gene_biotype = "protein_coding";
CD3E_XM_028834033	nimble	transcript	1	1258	.	+	.	gene_id "CD3E_XM_028834033"; gene_name "CD3E_XM_028834033"; transcript_id = "CD3E_XM_028834033"; gene_biotype = "protein_coding";

[Yenaled]

OK, the latest version does not support BAM files. The last version to support BAM files is version 0.48.0.

[bbimber]

OK, the latest version does not support BAM files. The last version to support BAM files is version 0.48.0.

Got it. I'm running 0.48.0 to try it.