The Institute of Clinical Sciences recently appointed Dr Giovanni Bottegoni. He joins the University from Sosei Heptares, bringing with him almost 15 years of academic and industrial experience.
Dr Giovanni Bottegoni
The Institute of Clinical Sciences recently appointed Dr Giovanni Bottegoni. Dr Bottegoni joins the University from international biopharmaceutical group Sosei Heptares, bringing with him almost 15 years of academic and industrial experience.
After several years working in different places and in different capacities, I started developing ideas that I wanted to explore further, pursuing an independent research program. The University had an opening in computational medicinal chemistry and what took my interest was:
I am looking forward to independently pursuing interesting ideas and the opportunity to build on the wide range of things that I’ve been exposed to in my career and blend them together in a clear research direction for the group. I would like to create a research group that does not exist 'in vacuo' but that is actually an organic component of the University’s ecosystem.
The difficult part of identifying the core research lines for the group won’t actually be separating bad ideas from the good ones; that will be easy. The hard part, and what I am looking forward to the most, will be prioritising the best ideas among many excellent ones.
Ideally, I can see the Computer-assisted Molecular Design (CAMD) group becoming the go-to place within the University for rationally designing compounds for challenging targets. My idea is to assemble a multi-disciplinary group of young and talented scientists with the aim of addressing, within my lab, the first early steps of drug discovery.
While I’ll start working mainly supervising computational projects, I would like for the CAMD group to eventually encompass expertise in synthetic chemistry and structural biophysics, to have in place the means for setting up a modelling – synthesis – testing positive feedback loop.
Predicting the future always is a tricky business. The risk of making very short-sighted predictions is high; even a genius like Lord Kelvin in 1895 went on records stating that heavier-than-air flight was impossible.
However, I will put forward a cautious idea quite well routed in the present: machine learning will profoundly change drug discovery the same way it has already changed retail, quality control, radiology, and so on. In a way, it is surprising that the drug discovery sector, which has been handling compounds by the millions for many years, that has been representing molecules as graphs for over two centuries, that pioneered the adoption of regression methods for molecular design, is not spearheading the adoption of machine learning on a global scale.
In any case, the revolution is here: there are already companies operating in the field of applying learning strategies to drug discovery and big pharmaceutical companies are hiring machine learning experts. When the hype will be over, when we will be past what Gartner calls the “peak of inflated expectations”, machine learning-based strategies won’t go away but will become ingrained in the drug discovery pipeline.
In my work, there is something that literally happens when you wake up in the morning. You often run very long simulations on a computer. These simulations can run for many hours, days even. After several years, I still feel that sort of curiosity when I wake up to quickly go and check the log file, to see what has happened during the night.
More in general, one of the best rewards, one of the things that truly motivates me, is when we spend quite a long time devising some specific compound based on models and calculations (and intuition!), and, later, when this compound is eventually synthesised and tested, its activity turns out to be exactly in line with the prediction.
