Slow loading for assembly-assembly alignment bams

[jrobinso]

From igv-help group...

Hi,
We are trying to visualize bams generated by minimap2, run as part of the dipcall pipeline (https://github.com/lh3/dipcall), for the Genome in a Bottle HG002 Q100 T2T assembly to GRCh38 with IGV. However, loading the tracks with these alignments when moving to a new position is really slow for both the desktop IGV app and IGV.js. We assume this is because they are chromosome scale alignments, links to the bams below. We would like to access the files from the FTP so that we can easily share the IGV session with collaborators. Do you have recommendations for IGV preferences or modifications to the bams, e.g. removing soft clipping or alternate alignments, to speed up loading these tracks? Alternatively, is there a better way to visualize assembly-assembly alignments in IGV?

Here are the links to the hap1 and hap2 bams
https://ftp-trace.ncbi.nlm.nih.gov/ReferenceSamples/giab/data/AshkenazimTrio/analysis/NIST_HG002_DraftBenchmark_defrabbV0.015-20240215/dipcall_output/GRCh38_HG2-T2TQ100-V1.0_dipcall-z2k.hap1.bam

https://ftp-trace.ncbi.nlm.nih.gov/ReferenceSamples/giab/data/AshkenazimTrio/analysis/NIST_HG002_DraftBenchmark_defrabbV0.015-20240215/dipcall_output/GRCh38_HG2-T2TQ100-V1.0_dipcall-z2k.hap2.bam

Thanks
Nate

[jrobinso]

You assumption is correct, the chromosome scale alignments are causing problems. Until there is a fix I suggest converting these BAM files to "bed" or "psl" format. You can convert to bed format using bedtools

[baozg]

Is it possbile to chunk the whole-genome alignment for IGV? Another alternatives would be chunking the genome and align them as reads #1551