Profile photo of ivan

by ivan

A talk by Dr. Skinner May 16th

May 14, 2014 in Labwide Announcements

Just a little reminder about the talk we will have on Friday.

It is by Dr. Michael Skinner, who's big claim to fame (?) is that non-genetic mechanisms can produce inheritable traits (i.e. chemical modifications that do not change the DNA sequence). I shall not pass any judgement (even though this disclaimer makes my bias obvious). The article attached tells a good story about his research, its implications and reception.

Please read this: Skinner Science

As one of the not-so-many evolutionary biology labs on campus, I believe that we should be able to ask a few provocative questions this Friday.

Profile photo of arnab

by arnab

Combined Annotation–Dependent Depletion

May 11, 2014 in Labwide Announcements

Following up,

I will be discussing the paper CADD published not so long ago in Nature Genetics, which implements a hyper-plane support vector machine trained to rank pathogenic variants in the human genome. The discussion will take place on friday, May16 from 3:00 pm to 4:30 pm


Profile photo of aaron

by aaron

Random Walks in Gene Space

May 11, 2014 in Labwide Announcements


I will be leading out in a discussion on Sahinalp's paper on random walks and driver genes, May 13 if anyone is around.

Here is the paper: HitNDdrive-Sahinalp-2014



Profile photo of shams

by shams

Polar Bear Genome

May 10, 2014 in Shams' Blog

Hey guys!

So it looks like they recently published a reference genome for a polar bear. There's a bit of an interesting discussion over the genes found to be under positive selection on the polar bear lineage

Here's the article:


Profile photo of Jason

by Jason

Something awesome

May 4, 2014 in Jason's Blog

This is hilarious. Thanks Ivan.


Profile photo of ivan

by ivan

A few curious papers.

April 3, 2014 in Ivan's Blog, Labwide Announcements

Here are a few intereseting recent development that have happened in the past week:

1) The paper I would like to talk about next Thursday:

2) A paper about "machine teaching":

3) A paper on differential gene expression detection:

Profile photo of Jason

by Jason

Detailed analysis of Neanderthal genomes

December 18, 2013 in Jason's Blog, Labwide Announcements

Svante Pääbo's lab has just published a detailed analysis of the newer high-resolution Neanderthal genome sequences! This is very exciting and worth a look; someone, maybe me, should present this paper in lab meeting in January.

See the paper here.

Profile photo of ivan

by ivan

A simple bridge between perl and C++

November 5, 2013 in Labwide Announcements

I have created a simple example script that wraps around the perlapi. This allows perl code to use C++ functions. There are a few accessory functions, that allow to convert a perl array into a c array and back. This may require a little cleanup so that it's not mandatory to de-reference the array in perl.

Hopefully the code is not too hard to understand. I would like to give a quick rundown of this on the Friday meeting if possible - it seems like a few people can benefit from this.

I will also write a short blogpost on my blog, and let you know when that's done.

The code is hosted here.

Profile photo of Jason

by Jason

Machine Learning Toolkit, Mahout

November 3, 2013 in Labwide Announcements

For anyone who's project or interests touch on machine learning, I came across the Apache project Mahout. It focuses on scalable implementations of machine learning algorithms, and appears to have a pretty active user community. It's written in Java.

The main project site is: Apache Mahout.

There is also an article about it in the current issue of Dr. Dobbs (the programming magazine).


Profile photo of ivan

by ivan

Short update on the Palantir visualization

September 26, 2013 in Ivan's Blog

I have put together a simplistic R script that draws a few circles next to the branches on a tree.

Screenshot - 09252013 - 06:27:27 PM
Thus far, it is a rather simple hack that uses a the obscure numbering of branches provided by APE. I foresee that it might be necessary to come up with a better way to encode branch identity. This really depends on what the output from Palantir is.

Another problem is that I am using the plotrix library for drawing things. Sadly, it does not scale well when a window is resized, so you might end up with elliptical circles. Additionally, I will have to play with the coordinates of the circle offsets to make sure they consistently fit on the picture.

Hopefully, when I get the data, it will not be too hard to make work properly.

The current usage information is provided in the repository description.



Skip to toolbar