Mailund on the Internet

As you might have heard, the PLoS journals have added access statistics to their papers, so you can see how often a given paper is accessed, either the html page or PDF downloads and whatnot.

Narcissistic as I am, I checked our CoalHMM PLoS Genetics paper.

I don’t know what I expected, but 15 thousand views is probably more than I did expect.  Not all of those are people who actually read the paper, of course, but the two thousand PDF downloads might be, and that is a lot of people…

I can just hope they liked it.

–

269-302=-33

We’re back again from Leipzig and had a wonderful time there.  It was really exciting to hear what they were working on there, and despite that we didn’t get in on any Neandertal analysis — the data doesn’t seem to be good enough yet that our method will give reasonable results — we did get in on a new genome project: the bonobo sequence project.  So all in all a successful trip.  Quite enjoyable as well, with all the very nice people we met and exchanged ideas with.

It didn’t bode well when we just started the trip.  The back end thingy of the car — that is a technical term — fell off just as we entered the autobahn, and that felt like a pretty bad omen.  We fixed it with some tape we bought at a gas station, though, and after that everything went smooth.

The place is really nice.  We stayed in rooms at the top floor of the institute.  Very nice rooms at that, with a nice balcony with a great view out of the town.  Of course, that meant that we didn’t see much of the town as there was little reason to even leave the building — and with a very packed program there wasn’t much time for it either, but with all the stuff going on it didn’t matter at all.  There wasn’t any time to get bored.

On the way back we needed to go over Copenhagen, as Mikkel had to give a talk there in the evening.  Well, about 100km outside of Copenhagen Mikkel wasn’t sure it was the right date for the talk, so we had to stop and check his papers, but it was the right date after all, so we went and he gave the talk.  That meant that the home trip was more than 1000km and we didn’t get back until late in the evening and I’m pretty beat now after too little sleep.

Since my brain has stopped working, I’m giving up on my Makefile hacking now and will go for a pint with some of the gang.  A perfect end to a perfect week.

–

268-301=-33

I’m using Make to implement a workflow, and here I use a bunch of files to indicate that certain tasks have completed.  They are not the results of the tasks, I just touch files to indicate that the given task is completed since that is the simplest way to tag that.

There is a bit of chaining going on, so a makefile like this is pretty representative:

all: foo.c

.PHONY: all

%.a:
	touch $@

%.b: %.a
	touch $@

%.c: %.b
	touch $@

Now, if I run this, this happens:

$ make
touch foo.a
touch foo.b
touch foo.c
rm foo.b foo.a

Why does it delete foo.a and foo.b?  I didn’t ask it to, and I certainly don’t want it to.  Tasks a and b are completed, and I don’t want to redo them later unless I explicitly ask for it, so why does Make delete the fact that the tasks are done?

It doesn’t delete the last file, foo.c, so it is not completely useless, just 99% useless…

–

268-300=-32

The paper has been available as “advanced access” for a while, but now it is out in its finished form: Ancestral population genomics: The coalescent hidden Markov model approach.

A short description, from the issue highlights:

Ancestral population genomics: The coalescent hidden Markov model approach, pp. 259–274

Julien Y. Dutheil, Ganesh Ganapathy, Asger Hobolth, Thomas Mailund, Marcy K. Uyenoyama and Marcy K. Schierup

The genealogy varies along the genome of species with close speciation times and large ancestral population sizes. Information on population genetics processes and the speciation process itself can therefore be extracted from genomic alignments, particularly when there is incomplete lineage sorting, i.e., part of the genome has a different tree than the species tree. These authors develop a framework for inferring ancestral population genetics parameters from an alignment of four species, using a hidden Markov model and coalescent arguments. They investigate the properties of the model by extensive simulations and apply it to human-chimp-gorilla-orangutan alignments.

–
265-299=-34

I’m off to Leipzig to visit Svante Pääbo’s group at the Max-Planck Institute there, to try to get in on future Neanderthal analysis.

The first Neanderthal paper should be out soon(ish) so of course it is too late to get in on that, but we hope to be able to use our CoalHMM methodology on future data.

I don’t know how much I will be able to blog while away, so you might not hear from me for a few days… we will see.

–