Possible Topics for Upcoming Meetings

[Lisa Rogers]

Hi guys,

I thought that I'd get the ball rolling on setting up possible topics for discussion over the next few weeks.  Maybe others can follow on and post a topic or two they can talk about or want to learn about.   

So for my project, I'll be getting back over 70 different strains of E. coli whole genome sequences.  I'll have to take the raw reads and assemble them into contigs.  And I had no idea how to do this before a few weeks ago.  So I got some sample data to practice with so I'd have some idea when I get my strains back.  So I thought that I could try and present some of the steps at an upcoming meeting and we can all learn together.  Or for the people who know how to do this they can offer me some hints and tricks.  I won't be free to talk about this till the 9th, as I won't be around on the 2nd.  So I can present at any session after the 2nd.   

Looking forward to seeing what other people want to discuss.

Lisa

[Kevin Rue]

HI all,

Thank you Lisa for offering those topics. Remind me after December, 2nd, and we'll make sure to assign some time to discuss and learn about whole genome assembly. It seems to me E. coli is a nice example to demonstrate the technique. Simpler than human I am sure!

I don't have any slide ready, but tomorrow I would be glad to do a live demonstration of the bioconductor CellMix package. This package takes genome-wide expression dataset, and uses cell-type-specific marker genes signature to deconvolute mixed cell-types samples and predict the proportions of each cell-types in each sample.

I will show you how the basic use of the package, and how I adapted it to deconvolute bovine samples (not supported by the package, details below.but spoiler alert for the meeting!).

See you tomorrow 25th November, 2pm, Conway Interactive Hub.

Cheers

Kevin

--------

A few more details about CellMix:

The package was developed by Renaud Gaujoux and supervised by Cathal Seoighe from NUI Galway. So far, the package contains only cell type signatures based on human and rat platforms, and largely focuses on whole blood and PBMC deconvolution. However, the package contains a converter enabling the application of this package to samples from mouse, drosophila, C. elegans, arabidopsis (thaliana), zerafish (Danio rerio), S. cerevisiae.

In addition, Renaud is open to including more deconvolution and pure cell-type expression signatures, if they prove good deconvolution accuracy.

Those of you not working with cattle may not have noticed, but cow is not listed in the comprehensive list above. Over the last few days, I mapped the human probesets used in one of their signature to bovine probesets (harder than it sounds). Now I have a fast procedure to deconvolute even species not supported by their package. Bovine deconvolution shows reasonable estimates of cell type proportions.

--
You received this message because you are subscribed to the Google Groups "OpenSequencing" group.
To unsubscribe from this group and stop receiving emails from it, send an email to opensequencin...@googlegroups.com.
Visit this group at http://groups.google.com/group/opensequencing.
For more options, visit https://groups.google.com/groups/opt_out.

--
Kévin RUE-ALBRECHT
Computational Infection Biology PhD Programme
University College Dublin
Ireland
http://fr.linkedin.com/pub/k%C3%A9vin-rue/28/a45/149/en