Extracting Reads from MEGAN

[GSviral]

Hi guys,

I have recently sequenced various gut metagenome samples, ran the contigs through BLAST, and created a taxonomic tree using the BLAST output.

After some preliminary analysis of known viruses I want to focus on the contigs which didn't result in any hits using BLAST.

To do this I wanted to extract the contig sequences under 'No Hits' of the tree. When I select 'Extract Reads' it simply extracts the contig name. For example the resulting FASTA file reads like this:

>contig00001>contig00002>contig00003>contig00004... etc.

There isn't any actual sequence data associated with these extracted reads.

Am I approaching this correctly? Is it possible for me to extract these sequences from MEGAN conveniently?

Thanks very much for any help.

[Romainb]

Hi,

Did you succeed to resolve the problem ? I am experiencing the exact same issue here... A mate in the lab uses the extract reads option successfully. We have the same version of MEGAN, and analyzing the same datas, but we can't figure it out.

Thanks in advance.

Rob

[GSviral]

Hey Romainb,

I was able to resolve the problem yup. What you have to do is:

When you import, for example, you BLAST output file (.xml or .tab etc.) you will have 3 boxes. Your input file will be your .tab for example. The third box is your output destination for your generated .rma MEGAN file.

The middle box accepts a FASTA file. What you can do is use the FASTA file that you used in your BLAST search and associate it with the MEGAN output. This will associate the actual contig sequences with the taxa assigned and you'll hopefully be able to successfully extract your reads of interest in a FASTA format.

Associating your contig FASTA file with your MEGAN output like this will also allow you to view alignments to reference genomes all within MEGAN.

I hope this helps!

[Romainb]

GSviral,

Thanks a lot, it totally worked !

We then realized that, if the fasta file and the blast file have the same name, MEGAN will automatically open the two files in their related boxes...