SpliceMap: New tool for detecting splice junctions from RNA-Seq data

[jm1234567890]

Hello SEQanswers Community,

Further Info: http://www-stat.stanford.edu/~kinfai/SpliceMap/

We at the Wong Lab have developed a new tool for the de novo detection of splice junctions from RNA-Seq data. It has been shown to be more sensitive than TopHat as of earlier this year and comparable in execution time (C++ version).

In it's current form it supports mapped reads from ELAND and SeqMap that are >=50nt long. The SeqMap program is included in the package, but ELAND is much faster for this purpose.

The C++ version is much much much faster. So try to use that one if you can.

Both the C++ and Python codes are designed for the Linux platform. If you use another platform leave us a note and we'll see what we can do. It seems to work on MacOSX if you compile everything from source (which will be up soon), this has not been extensively tested though.

Instructions are on the website, but if you are having any troubles don't hesitate to leave me a note here.

Please give it a try and let me know if you have any issues. We are actively developing this tool so we welcome all of your comments and concerns! The ease of use is very important to us, so let us know if anything annoys you.


John Mu

[jm1234567890]

Version 3.0 (C++) is released!

Please check the website: http://www-stat.stanford.edu/~kinfai/SpliceMap/

[Thomas Doktor]

I downloaded the C++ versions but the zip only contained executables, do you plan to release the source code or did I simply miss it on the webpage?

Do you plan to support SAM/BAM in a future release?

[kopi-o]

Any chance for color space support?

The manual is kind of sparse. It would be helpful if the home page would contain more information about the required input and so on.

[jm1234567890]

I downloaded the C++ versions but the zip only contained executables, do you plan to release the source code or did I simply miss it on the webpage?

Do you plan to support SAM/BAM in a future release?

Sorry about the source code. It will be on the website next week I think. I don't directly manage the website and I'll speak to my collaborator. It should be fine to use the code as long as you cite the paper.

Regarding SAM/BAM, the support will be added in future versions (maybe 1 month later). We are currently focusing on releasing all the features that were mentioned in the paper. The internal tools we use aren't really that user friendly haha

Thanks for your interest!

[jm1234567890]

Any chance for color space support?

The manual is kind of sparse. It would be helpful if the home page would contain more information about the required input and so on.

Yes, coloring of the splice junctions will be added in the next version. For example, novel junctions can be colored differently.

Yes... the manual is a bit sparse, when we have time. It will be gradually updated. I don't directly manage the website, so I can't personally change it. If you have any issues using SpliceMap, you can ask here or email us.

If you have any requests for features please let me know! Thanks!

EDIT: I think I misunderstand what you mean by color space. What do you mean exactly?

EDIT2: Ohhh. I understand now. Sorry I was not too familiar with that term. I will put it on the list of things to consider.

[kopi-o]

Thanks for the reply! Yes, I meant support for providing reads in SOLiD color-space format rather than FASTA reads from e g Illumina.

[phenotype]

Manuscript?

Hi,

Where is the manuscript that goes with the SpliceMap? I couldn't find it on the website.

Thanks.

[Thomas Doktor]

There is a link to the preprint on the frontpage, but you can grab it here: http://www-stat.stanford.edu/~kinfai.../SpliceMap.pdf

[UNCKidney]

Is it worth trying 37-bp SE reads... I have 7 lanes of data I could combine if that would make a difference

[jm1234567890]

Is it worth trying 37-bp SE reads... I have 7 lanes of data I could combine if that would make a difference

Hi, I'm sorry the shortest read length SpliceMap supports is 50bp. We could probably modify the program to go down to 45bp. But, anything shorter than that would not be useful.

SpliceMap is designed with the idea in mind to support longer read lengths. So unless there is huge interest in 45bp read lengths, we will keep the minimum at 50.

Thanks for your interest!

John Mu