What coverage is enough??

[valfdz]

Which coverage is enough to permit the assembly?

We are thinking sequence a bacteria and we want to know how many reads we will need to do.

We will use Illumina pair-end of 35 or 56 bp.

thx.

[Thorondor]

i would say the higher the better. ;-)

with 35bp more t hen with 56bp. and it also depends on your choice of the insert size. Since you do not expect many repetitive regions you could be good off with around 20x - 30x but more is better.

Do you have near relatives of the bacteria you want to sequence?

[valfdz]

Thank you.

I don´t know. I just know that is a cyanobacteria but i don´t know how near are from other cyanobacteria.

In another project we are thinking in sequencing a evolution strain of another cyanobacteria (this bacteria are sequenced). In this case with less coverage will be enough??

[Thorondor]

these are only guesses, keep that in mind!
with paired end and 56bp and a good reference you could go with less coverage maybe even with 5x, always depends on what you are looking for. since the cyanobacteria genome is small, why it is so important for you to know the minimum coverage you will need?

[valfdz]

Because i need to choose how many lanes i will use for sequencing.

[TonyBrooks]

I would say at least 25x, but more ideally. If you are mapping to a related genome, you might be able to get away with slightly less.

[Chipper]

Because i need to choose how many lanes i will use for sequencing.

How many samples are you planning to run per lane? 10 M 2X56 bp reads would give you more than 100x coverage for a 10 Mb genome, a single lane on HiSeq or GaIIx should give you many more reads.

[valfdz]

Today I was talking with the technician and She said me that thay are using TrueSeq protocol so she thinks that we will have about 30-40 Millions of reads per lane. In the worst case with 1 lane per sample is more than enough.

Thanks to everyone.

[BaCh]

Which coverage is enough to permit the assembly?
We are thinking sequence a bacteria and we want to know how many reads we will need to do.
We will use Illumina pair-end of 35 or 56 bp.

Errrrrmmmmm ... you do not want to use 36 mers for de-novo, this will be a total disaster, a cataclysm, a disaster, the apocalypse. At least if you expect to have somewhat presentable genome data.

Simply. Do. Not. Do. That!

Use reads as long as you can, 75mers at least, 100mers or more are better.

B.

PS: the above statement does not apply if you are just on a gene fishing expedition, then 36mers might be OK.

[valfdz]

Finally we will use 75bp pair-end so I hope we have enough data to do a good assemble.