Technical tips to plan a resequencing project

Question

[lessismore]

Dear all,
i'm dealing with deciding the sequencing technologies to adopt for a resequencing project of 1 specific plant variety. Genome size of this species is about 700Mbp.

1. Can you suggest me desired output, preferable (mixed?) sequencing technologies to use, and, if you know that, more or less how much will it cost me? Am i forgetting something? Should i know other things before making such a choice/ask for a quote to the company?
2. If you know that, can you suggest me the same things for a transcriptome sequencing?

If you tell me or point me to some website/review where i can look it up by myself is also very welcome. Thank you in advance for any tip!

[lieven.sterck]

one piece of critical info is missing: what is the purpose of the re-sequencing effort? SNP, SV , ... ?

[Joe]

^ I came here to ask this exact same question ;)

[lessismore]

Thanks for replying ! Ideally i'd like to have the highest recommended coverage to investigate SNPs and SVs in a species with a high level of heterozygosity.

[GenoMax]

Can you clarify if a good reference is already available? If there is no reference then you will first need to create one so it becomes a different problem. In any case the answer is almost always be short and long read combination.

[lessismore]

I confirm a good reference is available. Thanks!

Answer 1

[lieven.sterck]

Since you indicated you are interested in both SNPs and SVs, this might be a bit of a not so straightforward answer.

For SNP analysis it's advised to go for the high accurate technologies (Illumina, DNBseq), for SVs you are better of with the long read technologies as those give you better/more info on structural info.

You thus can notice that you're on both ends of the spectrum concerning seq-technologies. If money and time allows you could do thus both. If not , I think the best to go for is something like 150 nt PE reads on illumina, as this will give you both the accuracy you need for SNP as well as some longer range info for the SV.

Coverage wise, for SNP typically you would aim at 50-60x coverage to accurately call SNPs, for SVs with long reads you might go for less coverage.

[lessismore]

Dear, thanks for your very informative answer! I forgot to write that a good reference is available. Does your answer change in this case? I've seen that for example, for tea, coffee and sunflower the coverage was between x10 and x20
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6895910/pdf/genes-10-00929.pdf
https://link.springer.com/content/pdf/10.1007/s11103-020-00974-4.pdf
https://www.nature.com/articles/s41586-020-2467-6.pdf

[lieven.sterck]

yes, 10-20x might (/will) work as well ... even less in some cases ;-)

It all depends a little on the specific case (eg. polyploidy , allele frequency ... ) , in any case I would stick to a strict min of 20x , with 40-50x coverage you will be ready for all kind of scenarios/frequencies though. Moreover, sequencing nowadays is not that expensive any more, so in the end doing 20x or 40x will not differ that much.