Question: Ion Torrent data quality trimming
1
gravatar for vera.korneeva
24 months ago by
Russia
vera.korneeva10 wrote:

Hello!

I need to make contigs from the Ion Torrent data. This is a microbial genome fragment library. It supposed to be a set of 400 bp read. Due to quality reasons (I used FastQC to check) the reads are kind of good for the first 170-200bp. Our core sequencing facility releases the data as *.FASTQ files and the adaptors are supposed to be trimmed. I am not very experienced with NGS data analysis. The only program I used to clean my date was Trimmomatic but at that time I worked with Illumina reads.
Could you suggest me something like Trimmomatic for my Ion-torrent data, please?

Thanks for your tips.

ion-torrent trimming • 1.9k views
ADD COMMENTlink modified 24 months ago by 5heikki8.5k • written 24 months ago by vera.korneeva10

Since Ion data is in fastq format trimmomatic should work with it. What kind of cleaning are you looking to do? As I recall, Ion fastq scores are not totally equivalent to Illumina's. Perhaps you could try to use the data you have as is, with say SPAdes for assembly.

ADD REPLYlink modified 24 months ago • written 24 months ago by genomax71k

Thank you for your suggestion. The only cleaning I want to perform at the moment is to cut the low quality ends before assembly.

ADD REPLYlink written 24 months ago by vera.korneeva10
1
gravatar for 5heikki
24 months ago by
5heikki8.5k
Finland
5heikki8.5k wrote:

Optimal would be to use the ubam file straight from the sequencing machine with Spades assembler. You lose valuable information (that Spades utilizes) when you convert the ubam files to fastq.

ADD COMMENTlink modified 24 months ago • written 24 months ago by 5heikki8.5k

Are you sure about that? The Spades manual does not mention anything about it. It states that bam is allowed for Ion Torrent reads, but not that is preferred or in any way advantageous over fastq; just that for error-correction fasta should not be used (because the quality scores are lost).

ADD REPLYlink modified 24 months ago • written 24 months ago by Brian Bushnell16k

Yes, they use also flow information (obviously not present in fastq) for read error correction. I don't know why it's not in the manual but the main dev has at least communicated as much to me ~2 years ago. Edit. I read through my mail and came to realize that I did not discuss this specific topic with the dev, however, I'm convinced that I read about it somewhere. Also, if you dig through the source, you'll find many references to flow stuff under /src/projects/ionhammer

ADD REPLYlink modified 24 months ago • written 24 months ago by 5heikki8.5k

Thank you for your advice. Just to make things clear. Do you mean that I should better take completely row reads in bam format with all the adapters and barcodes in it and assemble it with Spades? Does Spades cut all the adaptors and barcodes itself? Sorry, if this is an obvious thing for you. There is nothing about it in the manual.

ADD REPLYlink written 24 months ago by vera.korneeva10

I'm talking about the files that are @ /var/www/output/Home at Torrent Server. They have been quality controlled and demultiplexed. These are the same ubam files that can be downloaded also through Torrent Server's web gui.

ADD REPLYlink written 24 months ago by 5heikki8.5k

Thank you, now everything is clear.

ADD REPLYlink written 24 months ago by vera.korneeva10
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1435 users visited in the last hour