Error "start too small" when running htseq-count on a sorted .bam file

Question

[elee228]

Hello,

This is my first time aligning scRNA-seq reads to a reference genome to analyze differential gene expression. I am using htseq-count to obtain count files for my different samples and I am receiving the following error:

Error occured when processing input (record #1507153 in file SRR6350437_STAR_Aligned.sortedByCoord.out.bam):
  start too small
  [Exception type: IndexError, raised in _HTSeq.pyx:376]

I can not seem to find anything online explaining this error...

This is the command I ran for htseq-count:

htseq-count -s no -r pos -f bam SRR6350437_STAR_Aligned.sortedByCoord.out.bam /Desktop/OFD_scRNAseq/Homo_sapiens.GRCh38.104.chr_patch_hapl_scaff.gtf > SRR6350437_counts.txt

Any help is greatly appreciated!

[swbarnes2]

I'm not sure that a reference that includes haplotypes and patches is best for STAR:

(emphasis original)

Generally, patches and alternative haplotypes should not be included in the genome

https://physiology.med.cornell.edu/faculty/skrabanek/lab/angsd/lecture_notes/STARmanual.pdf

I'd start by looking at what record #1507153 looks like.