Bacterial genome mate-pair advice

[Tom_C]

Hello everyone, I am quite new to NGS and would greatly appreciate some advice before I proceed with my sequencing project.
I am planning to sequence my pet bacteria's genome on an Illumina HiSeq using 100bp PE reads. The genome is ~4.7 Mb, and there are several reference genomes available. My question is should I also include a mate-pair library? And if so, how do you determine the fold coverage to use if I were to run both libraries on the same lane. I hope that question makes sense, I am still learning the basics if sequencing. Thanks!

Thomas

[mcnelson.phd]

If you're doing a single HiSeq lane, you'll have more than enough reads for a 4.7Mbp genome to split the lane 50%/50% using a standard TruSeq library and a Mate Pair library. It seems your goal is to use the standard library to map your reads to and find variants, but the mate pair reads will allow you to figure out if you have large chromosomal rearrangements, which I'd expect in a 4.7Mbp genome because of transposable elements.

Not to thrown an extra variable into play, but you might want to just do a Mate Pair library. You'll get essentially the same information as a standard library but with a larger pair distance which wouldn't really affect mapping but will drastically improve your ability to assemble de novo.

[MadsAlbertsen]

You would be perfectly fine with 5% of a HiSeq lane for a PE library and 1% for a MP library.

100x coverage should be fine. You can calculate the needed amount of sequence as:

100 x 4.7 mbp = 0.5 Gbp of data, which is 0.5/37.5 = 1.5% of a single HiSeq lane (assuming you run 2x100 bp).

I would use the TrueSeq PCR free kits to make a PE library for nice coverage of problematic regions - mostly high/low GC regions.

We normally do 500-1000 x coverage just to be sure we dont get any coverage drop out due to funky regions.

rgds
Mads