Profile cover photo
Profile photo
Guillaume Filion
Randomness is an attitude
Randomness is an attitude

Guillaume's posts

Post has attachment
A blog post where I explain my opinion about panel interviews in the academia (backed up by some data).

"I was very surprised that the top candidates often proved less productive than those we had ranked mediocre. How was this possible at all? Panels are unbiased, they combine multiple expertises, they allow for critical discussion, so they should be able to pick the best candidates... right?"

Post has attachment
"We learned a great deal about how to organize such peer coaching sessions in the academia and how to make the best of them, but this is not what this post is about. Instead, I would like to share more important lessons I have learned about working together and using the group as support and source of motivation."


Post has attachment
How to choose if your statistical tests should have one or two tails? This blog post in dialogue form helps understand some of the caveats.

"Imagine that I want to prove that male crocodiles are bigger than females. I set up a one-tailed t test and it turns out that females are a whooping thirty percent bigger than males. What is my conclusion?
— I guess you conclude that females are actually bigger.
— Wrong. I conclude that females have exactly the same size as males.
— But you just told me that they are bigger!
— Yes, and I also told you that I am doing a one-tailed t test.
— That does not make any sense."


Post has attachment
Here is a quick summary of our lab retreat. That was intense and very useful.

Post has attachment
We recently developed a technology to study the relationship between HIV and the genome of the host. With B-HIVE (Barcoded HIV Ensembles), we hope to understand why and how HIV sometimes becomes latent and hides in the genome of infected cells.

The gist of B-HIVE is to label viruses with a DNA barcode that allows us to trace them in a heterogeneous population. With this barcode we can tell which viruses are active, which are latent, and which are reactivated by different drugs.

To our surprise, we discovered that drugs have very different activities, depending on where HIV hides. We do not fully understand why that is the case, but this has important implications for the development of antiretroviral therapies.

Our work has just been published at Nature Structural and Molecular Biology.


Post has attachment

Post has attachment
A new wave of indexing algorithms based on the Burrows-Wheeler transformed changed the face of bioinformatics. For a long time I wanted to write a tutorial about these awesome methods. In this post I explain step by step how this all works to index the human genome. The detail is fairly technical, but this is a comprehensive resource for self-teaching.

Post has attachment
Here is an excerpt from the last post of Kevin Queral:

Then you’ll say, and you’ll be right : "How can we teach such a quality [empathy]? How a man can learn virtue ?" The answer is not a dodecagon, the answer is that we do not know how. At most, we have some leads.

We can not sell elixirs.

Post has attachment
How to write better peer-reviews? The great thing about publons is that we can now discuss about this openly and improve the system. I think that peer review should be a process of improvement more than selection. That said, I really like the suggestion of this blog post to send clearer signals to the editor in the review. Will do.

Post has attachment
This is many people's story.

"You went to high school and you learned genetics. You heard about a certain Gregor Mendel who crossed peas and came up with the idea that there is a dominant and a recessive allele. You did not particularly like the guy because there would always be a question about peas with recessive and dominant alleles at the exam. But you grew up, became wiser and just as you started to like him, you heard from someone that he faked his data."

More on this in the blog post.
Wait while more posts are being loaded