Check if nanopores reads is from the same species as the one used in an assembly

Question

[Olivier]

Hello,

I have wgs nanopore reads with only about 10x of coverage that has been generated from a given eukaryotic sample (as a single *.fastq.gz file).

Initially the goal was to produce an assembly but it's probably going to be hard with that amount of data.

In the meantime, I've found that an assembly has already been published and I was thinking to use it instead.

The issue is that my sample is from a cryptic species and I'm not sure that the species from the sample matches the one from the assembly.

Do you have an idea of how I could check if they match (or not) ? :)

[barslmn]

You can do mapping with bwa.

Answer 1

[colindaven]

10X is likely too low for assembly, but you can give shasta or flye a go and play with parameters (shasta is fast).

However, like you say, read mapping is going to get you further. You can map reads, as barsimn says, but use minimap2 not bwa as this aligner is more useful for long nanopore reads.

Watch out for supplementary and secondary alignments (see samtools flagstat on your bam file) and perhaps exclude these from downstream analysis.

Lastly, longshot or clair3 will allow you to call SNPs against the genome.

You can check your "cryptic species" status using alignments and number of SNPs to well characterized regions, eg 18S rRNA genes, single copy genes etc.

[Olivier]

Thank you for you complete response, I'll try that :)