Minimum coverage requirement for 50Mbp alignment?

[newkid]

As the title says? What's the minimum coverage required for alignment of approx ~50Mbp to a reference?

[byou678]

Could anyone give the general coverage requirement criteria for alignment?

Thanks.

[swbarnes2]

It depends what you are doing, and your question is so vague it's almost nonsensical.

To, say, call homozygous SNPs, 30x is a good goal. For RNAseq, you might want more.

[krobison]

To echo swbarnes2, you need to specify your problem both in terms of experiment design & goals.

In particular, what does the 50Mbp have to do with it? Is this the genome size? Do you have a set of clones which should add up to 50Mbp? Are you trying to capture 50Mbp? The latter two would require baking in some overage for vector or off-target sequences.

[newkid]

Looks like I have a lot to learn.. The problem is, I don't even know what questions I should be asking.

Is there a good beginner friendly guide out there that dives into practical application?

[Genohub]

Depth of Coverage Guide for Several NGS Applications

You can use the coverage requirements in this guide as a starting point. I also recommend taking a look at the ENCODE Consortium guidelines (referenced).

[Rocketknight]

In general, your total coverage target should depend on your experimental setup. As the other posters have mentioned, if you're sequencing a 50mb genome, then all of your reads should come from that genome. If you're doing 50mb exome sequencing from a much larger genome, then a lot of your reads will fall outside the target regions because the capture isn't 100% efficient, and you'll also have issues with some areas capturing or amplifying better than others.

Also, the coverage requirement will depend on what you want to get out of sequencing. If every sample is individually important (e.g. for clinical studies), and you want to see every SNP and indel, even heterozygous ones, in every sample then you'll need high coverage, like 100+ (say 3-4 samples per lane of a HiSeq 1000/2000, or a 1500/2500 running in high-output mode).

If you're more interested in population-level data where the exact genotype for a single sample isn't crucial, or you only care about homozygous mutations, you can probably go lower. If you're confused, give us some more info about what your goals for this sequencing project are and I'll try to suggest a reasonable coverage target.