Metagenomic read recruitment: best software

Question

[fhsantanna]

In your opinion, what is the best software to generate a figure of recruitment of metagenomic reads to a reference genome?

I have utilized Artemis, but the output was not proper for publication...

Any suggestions?

[Istvan Albert]

it is not clear what you mean by recruitment plot - there are many ways to show that

[fhsantanna]

I would like to do something like this:

Do you know a software that detects low read coverage regions in a reference genome?

[fhsantanna]

I found gnuplot as a possible solution. After using fr-hit for my analyses I got a table (out.frhit) as the following format:

read    size    evalue    pos1    pos2    hits    strand    identity    reference    rpos1    rpos2
IABX8XZ01DXRB3    46nt    6.8e-04    32    2    33    +    88.24    gi|239825584|ref|NC_012793.1|    851173    851206
IABX8XZ01CQO4L    231nt    1.0e-26    177    220    44    -    79.33    gi|239825584|ref|NC_012793.1|    977865    978043

Using the command in gnuplot

plot "out.frhit" using 10:8 with point #using the rpos1 as x and identity as y.

I got a plot that is kinda what I want. However I would prefer to use segments instead of points, the variables rpos1 as x1 and rpos2 as x2, for a given identity value (y). Do you know a style that could be used for it?

Thanks

[Jianshu_Zhao]

See here: https://github.com/jianshu93/RecruitmentPlot_blast

It was based on enveomics (https://github.com/lmrodriguezr/enveomics/blob/master/Docs/recplot2.md) packages developed by out lab. You can produce a bunch of plots for each of your reference genomes and use it for publication purposes.

For bam based recruitment plot please see here:

https://github.com/jianshu93/Competitive_mapping

It was based on this: https://github.com/KGerhardt/RecruitPlotEasy

We are going to have a new command line version very soon. I will update then.

Jianshu

Answer 1

[Josh Herr]

I'm not aware of any programs to automatically map metagenomic reads to a reference genome and then make a figure. You can do this by mapping your reads to your reference and then visualizing using many of the tools to do this (Tablet, Artemis, R packages). The figure you link to in your comment was probably generated by numerous tools and assembled in Illustrator or Inkscape.

As someone who works with metagenomic data, I would be very careful mapping metagenomic data to a reference. Is this a reference for a microbe in your sample that you have cultured and then sequenced a genome? What is your goal in doing this? (see Istvan's comment above -- what do you mean by recruitment plot in this context and what exactly are you looking to communicate?)

[fhsantanna]

I would like to map my metagenomic reads in reference genomes because there is a particular genus that is very frequent (more than 50%) in my samples. This way I would identify low coverage genomic regions, which could be genomic islands.

[Josh Herr]

The point I am trying to get at is understanding the genomic diversity at the level of genus or strain in your sample is difficult -- even if you have one genus which encompases 50% of your diversity (which I find hard to believe, but I don't know what your sample is).

[fhsantanna]

I work with thermophiles. I have sequenced thermophilic-enriched cultures derived from environmental samples. That is why there is a particular thermophilic genus that is abundant.