Greetings from Malaysia. I come in peace!

[Melissa]

Hi everyone,
I'm a PhD (Permanent head Damage) student from Malaysia. I'll be working on SNPs discovery via transcriptome sequencing using next generation technology. And btw, I'm a bioinformatics idiot and still figuring out how things work. Very soon I'll be dealing with overhelming of sequences data. Is there anyone out there who can make my life less miserable than it already is?


Cheers
Melissa

[zee]

I can, I used to work in a Malaysia. What platform are you using and which uni are you at?

[Melissa]

Hi zee, nice to meet you! I'm from UKM.
We have just decided to send some samples for Solexa mRNA-seq by the end of this year. Maybe will be using 454 for some other samples.

[zee]

Cool, we're also working with Solexa data from various samples. Have you decided on an analysis pipeline, alignment tool, etc?

[Melissa]

Not yet. Hopefully I will have a clearer picture when I attend Illumina Genome Analysis workshop at the end of this month.
feel free to email me at xxx

[apfejes]

Good luck with the mRNA-Seq. (We've been referring to it as Whole Transcriptome Shotgun Sequencing, or WTSS.) I've been working on this for close to a year now, so I might be able to help a little bit, if things get stuck for you.

[Melissa]

Thanks, Anthony! I have read your blog. It's great!
Particularly interested in the poster "protein coding sequence in follicular lymphoma". Especially SNPs detection in lymphoma patient transcriptome using 454 sequencing. Where can I find the journal article?

[apfejes]

Hi Melissa,

I'm glad to hear my blog has been useful - I'm certainly enjoying all of the conversations that start up from the comments!

At any rate, the poster you mentioned never made it to publication. After several months of working on it, our PCR verification of the SNPs came back and 99% were sequencing errors, rather than true SNPs. I've left the poster up because I think the methods might be of some use for people, despite the data being entirely useless.

It's also a great explanation of how I spent the first 5 months developing my own permanent head damage - there was much head beating on the wall after getting the sanger sequencing results back. (-;

[Melissa]

99%
Hmmm... what could possibly be the problem? From what I can understand from the poster, you used redundancy within the same individual to discover SNPs, is that right?
I am about to start my work on in silico SNPs detection using EST redundancy of the same individual. This is definitely a blow!

Can imagine your head becomes smoother shinier and your glasses become twice as thick.

[apfejes]

I think the issue mainly had to do with the poor quality of early 454 runs, my inexperience at filtering out highly redundant data, and my blind trust of the SNPs I was calling. I don't know how much of the blame goes to each of those, but I would like to believe it was poor quality data. (454-fans may claim otherwise.) Since it was done well over a year ago, I'd imagine that the same problems have been solved, and that we have a much better understanding of the biases in our samples, now. I don't think it'll be such a severe problem at this point. (I just came back from a talk by Dr. Jay Shendure where he said 90/95 SNPs found were successfully validated with a capture method and Illumina sequencing, which has similar biases.)

As for using ESTs, that's a slightly different case - it'll be hard to do the filtering that's most common- removing identical sequences with the same start and end positions. I'm not sure how that will work for you.

And yes, while my eyesight hasn't deteriorated, I definitely have a shorter haircut than I used to, and a permanent bruise on my forehead. (Periodically refreshed as I find bugs in my code.)

[Melissa]

Thanks for sharing.
We just got a very nice quote from GSC on the Solexa sequencing service. One of our collaborators has used the service and was very happy with his results.

[apfejes]

No problem. I hope your experiment goes smoothly! (=