bioinfo.se
Professor in bioinformatics, Stockholm University. Protein structure lover ( interactions predictions, evolution ..). Using machine learning as a part of AI for Sciences for halv my life. In addition to succén, I sometimes rants about sailing or skiing.
616 posts
1,617 followers
414 following
Prolific Poster
Conversation Starter
comment in response to
post
Yes, I was too early
comment in response to
post
Congratulations
comment in response to
post
I think we have to take into account the difference between fields. In our fields, the skills in data analysis clearly have a value (for now) outside academia. Plus, we are lucky that pharma is one of the few industries that appreciate/require PhDs. In other areas, it is very different.
comment in response to
post
But likely La Masia is significantly more successful (hard to find exact numbers). I think I'm around 10% succesrate if you count professors and 25% of you count so other academia as well.
comment in response to
post
Compared to fotboll we do quite well "less than 0.5% of the players aged under 9 that are signed by professional teams, go all the way to play for the first team at some point in their life." idsoccer.com/how-many-aca...
comment in response to
post
In a healthy (non-exponential growing system), each professor should educate one future professor (in their lifetime), so we should not be allowed to have more than two PhD students each (in total over our career) to guarantee a 50% success rate. Can we all agree to do this?
comment in response to
post
5. Wars are easy to start and hard to stop
comment in response to
post
We just changed d0 to be a constant of 5.0 (-d0 5.0 if you use USalign). But if you compare molecules of different sizes, you might need to run it twice (changing the order of query and target) or change a couple of lines in the code.
comment in response to
post
You have to not thar TM-scores (in my opinion) is not a good measure for NA structural similarity (it uses the same d0 normalization as die proteins which is inappropriate ). See www.biorxiv.org/content/10.1...
comment in response to
post
All this damage just to stop a deal between Iran and Trump (and win the next election) - it makes me so sad.