Mitochondria & plastid genomes

[bioman1]

Hi all,
I have illumina paired-end reads (WGS) from hiseq 2000. I want to map these reads to plant mitochondria and choloroplast genome and extract the reads for mitochondria and choloroplast de novo assembly.

1. I want download all plant mitochondria and choloroplast genome at one go. But when I go to ftp site it is difficult to get genomes especially to plant genomes. Do know how can I download all genomes at one go?

Below are the links to respective genomes:

For mitochondira genomes (Viridiplantae mitochondrial genomes - 94 records ): http://www.ncbi.nlm.nih.gov/genomes/...&opt=organelle

For choloroplast genomes (Viridiplantae plastid genomes - 455 records ): http://www.ncbi.nlm.nih.gov/genomes/...id&taxid=33090


2. Can I combine all the mitochondria genome using 'cat' command?. Can I index these combined genomes as reference mitochondrian genome and map my reads to this using bwa? or should I do map the reads to the individual genomes?

3. can I use unmapped reads as choloroplast reads? or should I need to again map the reads to choloroplast genome?

[Brian Bushnell]

To download all of them, in Linux, you can do something like this:

wget "http://www.site.com/stuff/*.fa"

Then:

cat mito*.fasta > dir/mitoCombined.fasta
cat chloro*.fasta > dir/chloroCombined.fasta
(you have to be careful that cat does not try to combine your output file with the input files).

Then:

I suggest using BBSplit for this kind of problem.

bbsplit.sh ref=dir/mitoCombined.fasta,dir/chloroCombined.fasta in1=reads1.fq in2=reads2.fq basename=out_%.fq outu=out_unmapped.fq -Xmx4g

(-Xmx4g should be adequate in this case but you may need to adjust it up or down based on available memory)

This will write reads to 3 files:
mitochondrial reads go to out_mitoCombined.fq
chloroplast reads go to out_chloroCombined.fq
unmapped reads go to out_unmapped.fq

These will all be interleaved, for paired reads. You can de-interleave them like this:
reformat.sh in=out_mitoCombined.fq out1=mito1.fq out2=mito2.fq

[maubp]

I think you can do this with NCBI Entrez which offers some quite advanced filtering, e.g. properties like completeness - see for example: http://blastedbio.blogspot.co.uk/201...-chimeras.html