Joan-Emma Shea - The Link Between Proteins, Alzheimer’s and Parkinson’s Diseases

This is what I was hoping to find when I went to the City of Cambridge library during Mental Health Week to learn about Parkinson's, Alzheimer's and other neuro-degenerative diseases, but all I could find was a large number of books that taught people how to live with and care for a person with Parkinson's or Alzheimer's and that was really, really depressing! See Toby on The Most Controversial Photograph in Biology.

15:17 About MDS (Molecular Dynamics Simulations). They model the action of the proteins as a fluid dynamics problem! That sounds crazy! But it's because they are trying to model the thermodynamical environment in which the protein folding occurs, and they are using a dissipative model, I think. They are studying Cryo-electron microscopy structures, which are the proteins in a minimum energy configurations, so they aren't wiggling as they do in vivo.

I recall that one of the goals of Robin Milner's development of the Pi Calculus was to be able to model systems like this. I wonder what sort of software they run on that server farm? See A. Regev, W. Silverman and E. Shapiro's Representation and simulation of biochemical processes using the pi-calculus process algebra in Pac Symp Biocomput (2001)

Despite the rapidly accumulating body of knowledge about protein networks, there is currently no convenient way of sharing and manipulation of such information. We suggest that a formal computer language for describing the biomolecular processes underlying protein networks is essential for rapid advancement in this field. We propose to model biomolecular processes by using the pi-Calculus, a process algebra, originally developed for describing computer processes. Our model for biochemical processes is mathematically well-defined, while remaining biologically faithful and transparent. It is amenable to computer simulation, analysis and formal verification. We have developed a computer simulation system, the PiFCP, for execution and analysis of pi-calculus programs. The system allows us to trace, debug and monitor the behavior of biochemical networks under various manipulations. We present a pi-calculus model for the RTK-MAPK signal transduction pathway, formally represent detailed molecular and biochemical information, and study it by various PiFCP simulations.

Subscribe to University of California TV.

It is not the case that everyone with a neuro-degenerative disease shows the symptoms:

Taylor Sullivan - She beat Alzheimer’s disease!? The Nun Study (summary)

Subscribe to Taylor Sullivan.

Early childhood influences were also studied.

UMN - Nuns help researchers discover more about Alzhimer's  (2010)

Subscribe to UMN Health.

This has been half-way quantified as "Cognitive Reserve"

Building Cognitive Reserve: The Key to Aging Brain Health

Subscribe to Buddhi Clinic.

See Three Meditations.

Google DeepMind Podcast - Hannah Fry interviewing Pushmeet Kohli AI: Supercharging Scientific Exploration

I don't really get this. They're talking about the structure of a given protein. But those CryoEM analyses show that this depends to some extent on the thermodynamic conditions. When you freeze the things they form crystalline structures that are surely different to their actual forms in living tissues. There is a whole field of dynamical protein behaviour that is clearly there but not studied. Maybe that's for the better though? It might be that the only purpose people have for scientific knowledge is make drugs and to try to engineer Natural systems. What I am asking is "What are those 1.8 Million people studying proteins planning to do with their knowledge?"

Subscribe to Google DeepMind.

I watched her talk again and I see that the analysis they are doing is just one, two, maybe three molecules in water at 300K (27 C), then heating up to 500K (230 C) or cooling it down to 290K (17 C) and then doing this over and over again, thousands of times to get a sample space which they can plot to see the distribution of certain lengths.

There's a bit more of my vague idea about using Pi Calculus here which I made a few hours later, on my way home: