Genometa-using reference-problems

[chayan]

1. The tutorial of Genometa said that 'After starting Genometa, the locations of the "Metatie-Fastalines" file and the "Lineage Mappings" file should be set in the preferences." These two files are also provided with the package, located in the "data subdirectory". Should i have to create these two in my way during the analysis (means these two can only be used only for the tutorial purposes), if so how to create those two files?? or i can use those two for my analysis as well? Secondly what is a GFF file and to create that from NCBI for my own work??

2. In the custom alignment settings option, whenever i set --trim3 50, the read count column always showing 0, no abundance is there...why this is happening all the time??

[colindaven]

1. These are supplied files which can be used for any analysis, provided you use the reference genome set provided. You do not have to create your own. Just check they have been set properly using the Preferences section in the GUI. If you do have names and mappings of names to lineages, this is probably all fine.
eg
ID: NC_002516
Name: Pseudomonas aeruginosa
Lineage: Gammaproteobacteria

2. Need more information. What is your
a) system - Linux?
b) aligner ? Bowtie ? BWA ?
c) read length.

Are you sure bowtie is installed properly ? Type bowtie on the command line to check this.

Hope that helps,
Colin

[chayan]

First of all thanks for your reply sir..

1. I have used "One genome per genus" reference genome set supplied with Genometa and used "bowtie-build -f allgenomes.....per_genus.fa refenence" to generate to "reference.1.ebwt" (along with 5 other files), which i finally used as my reference set as mentioned in the tutorial. Set all the paths, preferences as suggested. But still when used {path/to/bowtie -t path/to/reference --sam -p 15 -n 3 -l 40 -e 200 --best --trim3 140 -q path/to/dataset.fastq path/to/output/dir/output.sam in the "custom alignment setting"} no abundance is generated although other colums are seems to be fine, no lineage also shown to be present in the .csv output.

2. a.Linux ubuntu 12.04 LTS 64 bit
b. aligner: Bowtie

$ bowtie
output: No index, query, or output file specified!
Usage:
bowtie [options]* <ebwt> {-1 <m1> -2 <m2> | --12 <r> | <s>} [<hit>]
with all the options-

seems to be fine???

c. Read length- IonTorrent with avg read length of 193bp.. here i want to use 50 bases from the beginning of the read (5' end)..

Whats going wrong??

[chayan]

Can i use use fastx-clipper from fastx-toolkit here to trim the reads and then use bowtie??? Is that will be going to affect my analysis by any means???

[colindaven]

Hi,

if you have set all the paths and bowtie runs then congratulations, you're making progress.

path/to/bowtie -t path/to/reference --sam -p 15 -n 3 -l 40 -e 200 --best --trim3 140 -q path/to/dataset.fastq path/to/output/dir/output.sam in the "custom alignment setting

These settings look fine.

Of course you can use any trimmer to trim the reads down to a fixed size or by quality. Just keep in mind bowtie1 works best with short reads below about 70 bp in my experience.

You can also try running the command generated on the linux command line and seeing if a SAM file is generated, then reading that into Genometa.

[chayan]

ok thanks..yes that i know and that's why i trimmed the reads to a length of 50 bases..

and yes i tried that and it generated the SAM file..after changing it to BAM format (which GENOMETA automatically perform) results are visible but, as I mentioned earlier, with no read count..somewhere i read that it may not work properly in an insufficient memory usage machine, but mine is quad core 64 bit machine with 20 GB of ram and >500gb of free space...so this cant be a problem..still did not get any idea why this is happening??

[chayan]

Should i reinstall bowtie and Genometa from a very beginning?? or any other dependencies needed for the proper functioning?

[colindaven]

Have a direct look at the SAM file. It looks like Bowtie did not run successfully, or that the reads are not being parsed successfully from the BAM.

You could try:
bowtie --version
bowtie version 0.12.7

You can also check that alignments do exist in your bam file with samtools:
samtools flagstat x.bam
samtools idxstats x .bam

Also check for alignments in the SAM

grep "NC_" x.sam

[ret4andry]

Is that will be going to affect my analysis by any means?