Image for post
Image for post

Ok. So 2020 wasn’t what we all expected it to be….

Many of my new year's resolutions/goals from last year involved traveling to different countries to spend time in nature and attending my favorite conferences in person. I couldn’t make those happen, but I’m still happy with the way that 2020 turned out. Instead of looking outwards, I took a lot more time working to work on my own personal growth and development as opposed to focusing on optimizing external metrics.

Top reflection topics from 2020

  1. Consistency over intensity — How to set better goals

Image for post
Image for post

2020 has objectively been a hard time for science, with the rearrangement of project priorities and funding and the lack of serendipitous in-person conversations at conferences. We won’t have pictures like those taken at the Solvay conference, scientists shoulder to shoulder. However, we do have Zoom screenshots, and that’ll do :)

I recently had the amazing opportunity to attend the Harvard-Stanford Symposium on Drug discovery! Topics ranged from domain extrapolation in chemical spaces to protein-protein interaction (PPI) prediction studies and design of COVID diagnostics. Below are the summaries of the individual talks.

Regina Barzilay — MIT

Professor Barzilay talked about how there are four components to improving the use of AI in drug…

An attempt at answering humanity’s (arguably) most important question through the lens of a 17th-century philosopher

Image for post
Image for post

An Introduction to the Classic Descartes Story

What if everything that you’re seeing in front of you is completely false? Sounds absurd, but the truth is, there’s no way to deny it. You could be dreaming, or maybe we’re in some sort of simulation where we’re being programmed to think in certain ways. Again — these thoughts are quite absurd. But, there’s no way to absolutely prove or disprove these ideas, and that’s why we can’t put them aside. This is the way that René Descartes, a 17th-century philosopher, mathematician, and natural scientists started questioning everything that he thought or believed. He saw his beliefs and logical views of the world as a bag of apples, where one rotten apple (false belief) could contaminate the rest. …

On Eternal Recurrence, Dimorphism, and Superhumans

Image for post
Image for post

When you think about people in history who went against the rest of society to conceive of an ultimate truth, what often comes to mind? Most people might of think of Galilelo Galilei or Nicolas Copernicus, who were declared heretics for promoting the heliocentric model, a model that seemed to directly contradict pre-existing Christian doctrines.

One name that might not immediately come to mind for most is Friedrich Nietzche; nevertheless, he most clearly defied societal norms. …

Why is it that we still haven’t been able to figure out how to cure diseases like Alzheimer’s? It’s mainly because:

  1. We don’t understand the main mechanism of the disease (the pathways involved in misfolded amyloid accumulation). That means that when we choose a target, we’re not entirely sure if it is ‘druggable’ — that is — whether it will have the effect that we intend it to have.

We’ll see more generally how machine learning is allowing us to get a better shot at more broadly solving both of these problems. …

Stealing the most rewards 👀🏆 using reinforcement learning

Image for post
Image for post

In this article, I go over the basics of reinforcement learning and the BANDIT algorithm along with explained code implementations. Skip the first section if you’d like to jump to the code implementations.


Imagine that you’re playing games at a casino and you have multiple slot machines that you can hedge your bets on — say 10 machines.

How do you know which machines to start with? And how do you know which machines to select after if your objective is to maximize reward?

This is the commonly stated multi-armed BANDIT problem within Reinforcement learning.

Reinforcement learning itself is commonly about having an agent that interacts with its environment. By interacting with its environment, it receives feedback, which it can then use in the next time step to better inform it’s decisions. …

Get an Intuition for the Difference between SGD vs RMSprop vs Adam optimizers

Image for post
Image for post
Smooth loss landscape visualization. Credit —

When people first start out machine learning in PyTorch, you might see a PyTorch script like this.

Let’s decipher the black box — A Simple Explanation to Force Field functions like CHARMM, programs like Rosetta, and cutting-edge Deep Learning Algorithms

Image for post
Image for post

If you thought about each of our cells like a miniature city, our proteins would play every role from transportation to housing, from banking to agriculture. They have complex three-dimensional structures and they do pretty much everything in our cells. I’d argue that our proteins should receive more attention than our genome because our genome is ironically the simplest parts of who we here. As information propagates through the transcriptome (RNA) to the proteome, our cells get more and more complex, and that’s why protein structures are crucial to try and decipher this complexity.

Of course — one of the best ways to understand protein function is through the structure. Yet this protein folding problem is something that we’ve been struggling to solve for decades. If you allowed a protein with an average number of amino acid residues to fold, the protein would take an eternity (no exaggerations) to sample all possible conformations (If you’re more interested in this, check out something called Levinthal’s paradox (it essentially states that there are an astronomical number of protein conformations)). It’s like having a super complex math problem and trying all possible values on a continuous spectrum from one to infinity, hoping to eventually land somewhere. …

Takeaways from Vishen Lakhiani’s “A Code of the Extraordinary Mind” and other books that can change your life.

Image for post
Image for post

What if I told you that you aren’t the one making your own decisions.

Ok, now there’s a whole argument concerning determinism and randomness along with the whole metaphysical and quantum definitions of consciousness, but aside from all of that — what if you (I mean the “you” that you perceive as ‘you’) weren’t the one making your decisions.

That initially wouldn’t make sense. …

An overview of some current methods to take a protein’s sequence — and get it’s three dimensional structure

Image for post
Image for post

Proteins do pretty much everything in our cells, from moderating the regulatory mechanisms in our cells to catalyzing other very important metabolic and cellular reactions. Therefore, acquiring structures for these proteins is super important. However, conventional mechanisms for protein structure determination are extremely time and cost intensive. X-ray diffraction, the more traditional method, can take weeks to months to acquire a high resolution structure due to the inherent inefficiencies in the scientific procedures used to prepare proteins for crystallization.

In the past decade, we’ve seen an exponential increase in the number of computational tools, the most prominent of which is Rosetta, a biophysical modeling tools that can do ab initio protein folding simulations (that is, getting a 3 dimensional outputted protein file with sequence input). These kinds of computational tools are extremely valuable because they can help us to simulate and acquire protein structures without all of the resources involved in other more traditional physical methods. …


Mukundh Murthy

Innovator passionate about the intersection between structural biology, machine learning, and chemiinformatics. Currently @ 99andbeyond.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store