SEQanswers

SEQanswers (http://seqanswers.com/forums/index.php)
-   Bioinformatics (http://seqanswers.com/forums/forumdisplay.php?f=18)
-   -   How to call the core genome of bacteria (http://seqanswers.com/forums/showthread.php?t=32517)

NGS group 08-05-2013 07:20 AM

How to call the core genome of bacteria
 
Hi everyone,

I am trying to build a SNP tree from a set of bacterial genomes (the same species), and the first step is going to get the core genome. However, I did not find any detail workflow for that. Does anyone know some software or pipelines that can complete such task?

Thank you very much!

Victor

rhinoceros 08-05-2013 10:09 AM

You could look around NCBI's ftp. Maybe your species is already in the SNP DB? If not, you might have to do all-vs-all blasts and clustering (maybe OrthoMCL) to find out the shared genes. What you want might be possible with eutils too. Oh and there's this pipeline called HAL that might do what you want..

themerlin 08-05-2013 11:39 AM

How many genomes are you working with? And are you working with assemblies or raw reads?

For a smaller number of genome assemblies, you could do whole genome alignments with Mugsy (http://sourceforge.net/projects/mugsy/files/) or ProgressiveMauve (http://gel.ahabs.wisc.edu/mauve/), filter for blocks that are common to all genomes, then infer a phylogeny from the concatenated alignment? Is that what you had in mind?

Jason

NGS group 08-05-2013 01:01 PM

Hi rhinoceros,

Thank you for your reply. The aim of my project is to distinguish outbreak strains, so I have to do the analysis by myself. The HAL pipeline is what I want exactly. However I am not so familiar with commend line, is there any other more user-freindly pipeline or programs under Windows system?

Victor
Quote:

Originally Posted by rhinoceros (Post 112584)
You could look around NCBI's ftp. Maybe your species is already in the SNP DB? If not, you might have to do all-vs-all blasts and clustering (maybe OrthoMCL) to find out the shared genes. What you want might be possible with eutils too. Oh and there's this pipeline called HAL that might do what you want..


NGS group 08-05-2013 01:11 PM

Hi Jason,

Thank you for your reply. I am working with 15-20 genomes, which have been assembled. I am using Mauve now, but I think it is not easy to filter the common sequences by manual in Mauve. Do you mean that I just concatenate all common sequences by copy paste? Which alignment and phylogenetic tools are you going to use afterwards? Also Mauve?

Victor

Quote:

Originally Posted by themerlin (Post 112590)
How many genomes are you working with? And are you working with assemblies or raw reads?

For a smaller number of genome assemblies, you could do whole genome alignments with Mugsy (http://sourceforge.net/projects/mugsy/files/) or ProgressiveMauve (http://gel.ahabs.wisc.edu/mauve/), filter for blocks that are common to all genomes, then infer a phylogeny from the concatenated alignment? Is that what you had in mind?

Jason


themerlin 08-06-2013 07:39 AM

Victor,

If you have an aversion to command line tools, you could try this web server:

http://cge.cbs.dtu.dk/services/snpTree/

You can upload assemblies, fastqs, vcfs, and/or bams and it will compute a tree. I haven't used it, but it might fit your needs.

Jason

NGS group 08-06-2013 08:13 AM

This is a nice web tool. Thank you very much, Jason.

Quote:

Originally Posted by themerlin (Post 112688)
Victor,

If you have an aversion to command line tools, you could try this web server:

http://cge.cbs.dtu.dk/services/snpTree/

You can upload assemblies, fastqs, vcfs, and/or bams and it will compute a tree. I haven't used it, but it might fit your needs.

Jason



All times are GMT -8. The time now is 06:10 AM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2019, vBulletin Solutions, Inc.