I’m Matteo Ferla, a computational biochemist who dabbles heavily in cheminformatics.
I am half-Italian half-English and to further confuse things I've lived in a few places:
I did my undergrad at the University of Bath (Somerset, UK) with placement at Dartmouth College (NH, USA),
I then lived in London, then did my PhD at the University of Otago (NZ), worked at Biosyntia (Copenhagen, Denmark)
and then came to Oxford.
My PhD was on enzyme evolution, specifically on enzymes where the multitasking enzymes were favoured
due to the fitness cost of having too DNA,
such as in the bacterium Pelagibacter ubique, which lives in the phosphate poor open ocean:
my favourite analogy for these enzymes was a spork, a tool that saves weight but is more awkward than a fork or spoon (i.e. lower catalytic rate) and more prone to accidents (ie. more side products due to its enzymatic promiscuity). The interest in these was because during early organisms in the primordial soup would have had to be very crude sporks before the development of chromosomes. Although, I will admit most interest came from the fact my thesis had
an infographic as the acknowledgements.
I then worked further on genome mining bacteria and archaea for fun enzymes: there are many and evolution has no qualms at implementing sloppy solutions. My favourite enzyme however was a sensible Rossmann fold enzyme I dubbed internally as bioÆ —I liked the name: Danish orthography is confusing (cf. Kamelåså on YouTube) but it was pronounced as written.
I moved to Oxford at first to evolve a protein system, DogCatcher,
which could be considered a suicide enzyme without any small molecule substrates —it was odd not having small molecules around, but the transition state was sufficiently weird to parameterise than it made up for it.
I like molecular thermodynamics, but I quickly found that the biggest challenge was not poor Gibbs free energy differences, but with solubility and aggregation
—protein are not isolated spherical cows in a vacuum it seems.
After that in the Oxford BRC genomics theme, I modelled protein variants that caused diseases,
namely explaining amino acid that were the opposite of what one would want to engineer.
There too molecular thermodynamics was not king,
but other things such as differential disfavouring of alternate conformations,
deregulation from loss of post-translational modifications,
sequestration of binding protein and so forth kept playing a role,
especially in gain of function variants. To help others untangle these puzzling variants,
I wrote Venus,
a tool to collate the information needed for this, and Michelaɴɢʟo a tool to help share protein structures.
During the Cᴏᴠɪᴅ19 pandemic I worked on a Fragmenstein,
a tool to place or merge molecules faithfully to the 3D conformations of the inspiration hits of a fragment screen despite making monsters that need a wee jolt of energy removed from them to be feasible. Now I am in OPIG and XChem (Frank von Delft) were I am continuing this work and expand in new directions.
After my PhD I ditched Perl and Matlab (please do not ask me my age...) for Python.
As a result I am a die-hard loyalist of it!
Ironically my first transition project was
a Dungeons and Dragons encounter simulator,
which I then made a web app for (hosted on an Intel Nuc sitting under my router at home).
This led to a few other web apps but academic in nature (such as
an directed evolution toolkit,
an epistasis calculator and
https://michelanglo, a tool for protein dissemination),
however I never seem to get as much traffic despite these being much better written.
I am an avid PyRosetta user, and frequently blog tutorials on it and other things on my blog.
I like doing so as it feels rewarding making my time spent figuring something out helpful to others
—answering an thankful email from a stranger asking for a clarification or two is especially rewarding even if my time these days seems run away from me.