View Single Post
Old 02-04-2018, 10:36 PM   #4
David Eccles (gringer)
Location: Wellington, New Zealand

Join Date: May 2011
Posts: 836

I can tell you what I use, but it's unlikely to be what other people prefer, because there's a lot of manual computational legwork involved.

De-novo Assembly -- Canu. Flye and MARVEL sound promising, but I've yet to test them out.

Alignment -- minimap2 and/or LAST, depending on the application. LAST is particularly good at quickly picking up duplicated adapters in chimeric reads; I combine the output of LAST with my work-in-progress maf_bcsplit script to make results a bit easier to understand (it parses the MAF format produced by LAST into one-line-per match statistics like start/end/mapped percent).

Differential gene expression -- count transcripts directly via mapping with minimap2, use DESeq2 for differential expression tests. Kallisto/Sleuth may work as well, but possibly not given that I didn't have great results when using Salmon for gene counting based on minimap2-mapped transcripts.

SNP/INDEL detection -- I currently use samtools mpileup in combination with my own readstomper script to get base-level genotype frequencies. I'd like to look into doing a read-level haplotyping analysis, or at least multi-genotype (e.g. see this paper, or this paper), but have yet to try that. For mapping to known strains / genomes, Centrifuge should work; I think it's what ONT uses now for their What's In My Pot analysis tool.

Last edited by gringer; 02-05-2018 at 01:04 AM.
gringer is offline   Reply With Quote