Unrecognised chromosome value (e.g. chr16_KI270853v1_alt) in the UCSC Genome Browser

Question

[ManuelDB]

When creating a bed file in UCSC Genome Browser from a list of gene names. I have noticed strange chromosome names

chr16_KI270853v1_alt 1481239 1481396 NM_022844_exon_39_0_chr16_KI270853v1_alt_1481240_r 0 -

chr16_KI270853v1_alt 1495896 1496258 NM_022844_exon_40_0_chr16_KI270853v1_alt_1495897_r 0 -

chr16_KI270853v1_alt 1514933 1515023 NM_022844_exon_41_0_chr16_KI270853v1_alt_1514934_r 0 -

What is this? I have seen how I have all my exons correctly with the correct chromosome value (e.g. chr1 or chr2) but at the end of the file a small proportion of these appears. These unrecognize (for me) values only appear when working with assembly hg38 but when repeating same operation with hg19, these rare values are not in the output file.

[Pierre Lindenbaum]

UCSC: FAQ: https://genome.ucsc.edu/FAQ/FAQdownloads.html

https://genome.ucsc.edu/FAQ/FAQdownloads.html#downloadAlt

The chr_alt chromosomes, such as chr5_KI270794v1_alt, are alternative sequences that differ from the reference genome currently available for a few assemblies including danRer11, mm10, hg19, and hg38. These are regions of the genome that exhibit sufficient variability to prevent adequate representation by a single sequence. UCSC labels these haplotype sequences by appending "_alt" to their names. etc...

[ManuelDB]

Thanks for your reply.

I have another question. I have created a bed file introducing a list of gene names in UCSC Genome Browser to target later a variant calling application. Can I delete this list of alternative sequences? If I do, am I not covering all my target genes? If I run my variant calling program, it returns a fail because tchr_alt chromosomes are not recognised.