Facing Error While Providing Input To Megan

Question

[vijay]

Dear All,

I am trying to analyse a set of metagenome sequences using MEGAN upon which I am facing some issues. I created a local BLAST database in my system that contains 16S rRNA sequences. I ran a blastn against this database and got the output. When I tried to input the file into MEGAN through the import from blast option, I encounter the below error.

java.io.IOException: Line 1: ';' expected, got: 's'

Meanwhile I tried running a blastx against the nr database for the same set of sequences, which I was able to analyse through MEGAN successfully .I want to understand whats actually causing this error. Is this a problem with the local database that I have created and is there something else that needs to be addressed??

Kindly help me out with this.

-vijay

Answer 1

[Josh Herr]

Hi Vijay, It sounds like something is wrong with your query and/or database. MEGAN is expecting a semicolon to separate part of your matrix, but it appears to be missing.

First I would visually inspect your local 16S sequence database (you didn't tell us which one you are using: Silva, RDP, greengenes, etc?) and make sure you imported it and unpackaged it correctly. You should be able to easily do this at the command line or in a text editor. Next, do the same with your BLAST output and make sure it is in the correct format that MEGAN accepts. Since the error is occurring in your first line, in my opinion it's probably a parsing error or something related to your BLAST output, and not a one off error or one of your sequences.

In the last two questions you asked (Importing Blast File Into Megan and Loading Input File In Megan), Istvan and I mentioned (and salah refers to), and if you read the MEGAN manual, you will find that you have to provide BLAST accessions which correspond to MEGANs pre-loaded NCBI taxonomy accessions. If your BLAST format is not corresponding to MEGAN you need to make sure it's formated correctly, as MEGAN cannot correctly correspond your BLAST to the phylogenetic position of your reads based on the NCBI phylogeny.

Answer 2

[salah.edinne]

If you are using blast output format 8 i.e. tabular then you should select the correct GI accession numbers to NCBI taxon ids table i.e. nucleic NOT protein.

http://ab.inf.uni-tuebingen.de/data/software/megan4/download/welcome.html

Best,