Question: Assembly of whole metagenomics data
0
gravatar for kalsikamal10
4.5 years ago by
kalsikamal100 wrote:

Hi all,

I want to know about the best assembly tool for whole metagenomic illumina paired end data of 101bp reads and what are the parameters to check best assembly for metagenomics data.

assembly • 2.6k views
ADD COMMENTlink modified 4.5 years ago by Skeletor90 • written 4.5 years ago by kalsikamal100

I used SOAPdenovo for metagenomic assembly but never found the good result. I tried it on the a set of HMP samples to reassemble them but i could not get as good assembly comapred to HMP assemblies. How is it work with you ? 

ADD REPLYlink written 4.5 years ago by kalsikamal100
4
gravatar for Mikael Huss
4.5 years ago by
Mikael Huss4.6k
Stockholm
Mikael Huss4.6k wrote:

I've found that Ray (Ray Meta to be specific) works well for me in this type of scenario. Also IDBA-UD and SPAdes have given consistently good results.

ADD COMMENTlink written 4.5 years ago by Mikael Huss4.6k
1
gravatar for iraun
4.5 years ago by
iraun3.5k
Norway
iraun3.5k wrote:

If you read this thread Denovo Assembly Of Paired And Mate Paired Reads , maybe you can get an idea about the starting point ... I suggest you to start with SOAPdenovo, I'm working in a metagenomics pipe and after reading different papers in the literature this is the tool I've chosen.

Hope it helps.

ADD COMMENTlink written 4.5 years ago by iraun3.5k

I used SOAPdenovo for metagenomic assembly but never found the good result. I tried it on the a set of HMP samples to reassemble them but i could not get as good assembly comapred to HMP assemblies. How is it work with you ? 

ADD REPLYlink written 4.5 years ago by kalsikamal100
0
gravatar for 5heikki
4.5 years ago by
5heikki8.4k
Finland
5heikki8.4k wrote:

In our case, Meta-IDBA worked relatively well with LR-trimmed 100 bp pe reads. I recall k-range was something like 20-75.

ADD COMMENTlink written 4.5 years ago by 5heikki8.4k
0
gravatar for Skeletor
4.5 years ago by
Skeletor90
Calgary, Alberta Canada
Skeletor90 wrote:

I would recommend playing around with some simulated data using a tool such as metasim in order to benchmark some of the assemblers out there:

 http://ab.inf.uni-tuebingen.de/software/metasim/

 

There is a tool called metaquast that will help you evaluate your assemblies:

http://bioinf.spbau.ru/en/content/quast-22-released

 

Unfortunately it is hard to evaluate metagenome assemblies on real world data as we don't know what should be in our samples. Metrics such as the N50 size aren't meaningful if your assembly isn't accurate. This is especially problematic for metagenomes as you are dealing with multiple organisms in your sample. Parameter optimization is also tricky due to having multiple genomes with different coverage levels. For example, you might find that using a certain k-mer value is good for the genomes with higher coverage but not those with lower.

 

 

 

ADD COMMENTlink written 4.5 years ago by Skeletor90
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 723 users visited in the last hour