Exploring the SARS-CoV-2 Main Protease (MPro) Structures

Like many other groups around the world, we've been doing some virtual screening on the SARS Cov-2 Main Protease (Mpro).  There are now a lot of crystal structures of Mpro available.  How do we decide which structures to use for docking?  We'd like to select a diverse set of protein conformations that will enable us to explore multiple binding site interactions.  In this post, I'll provide an overview of a Jupyter notebook that my colleague Nic Pabon put together to explore the Mpro crystal structures.  While this notebook is oriented toward Mpro, the techniques discussed here can be applied to any protein.

In this notebook, Nic used the open-source, Prody toolkit to perform a number of analyses on the Mpro fragment structures recently released by the team at the Diamond Light Source.  The notebook begins with an overview of some of the basic capabilities available in Prody.

  • Reading PDB files
  • Properties (number of atoms, residues, chains)
  • Per atom properties
  • Hierarchical indexing
  • Extracting coordinates
  • Extracting backbone coordinates

In the next part of the notebook, we explore the flexibility of Mpro and use the root mean squared fluctuation (RMSF)  to identify flexible residues.   We can then compare RMSF for the structures with covalent and non-covalent ligands bound.



We can also use Prody to put the RMSF values into each atom's temperature facture field.  After this, we can color the structure by the temperature factor to indicate the more flexible regions of the protein.


One common task we carry out prior to performing a virtual screen is identifying a representative set. of binding sites for docking.  One way to do this is to cluster the structures and select a representative from each cluster.  In this example, we first calculate the root mean squared deviation between all pairs of structures. This creates a vector for each protein that contains the RMSD to each other proteins.  We then use the k-means algorithm, as implemented in scikit-learn to cluster these vectors into a predefined number of clusters.

One interesting aspect of this notebook is an interactive viewer that enables the user to examine and compare the clusters.  The animation below shows how this works.


In addition to clustering, we can also use Principal Component Analysis (PCA) to compare the structures.  In PCA, we project the binding site coordinates into two dimensions.  In this 2D plot, related structures will be close together, while binding site conformations that are different will be farther apart.  In the plot below, each point representing a protein structure is colored by cluster id, with the cluster centers represented by "X"s. 



Finally, as I mentioned above, there are new structures of Mpro being released every day.  One way to find these new structures is to perform a BLAST search.  In this notebook, Nic shows how you can use Prody to convert an Mpro structure to a sequence.  This sequence can then be used to perform the BLAST search for structural homologs.  The homolog structures can then be aligned and used for subsequent analysis an virtual screening.

As usual, the code for this post is available on GitHub.  Please leave a comment if you have questions or improvements to the code. 


Comments

Post a Comment

Popular posts from this blog

AI in Drug Discovery 2023 - A Highly Opinionated Literature Review (Part I)

Image
Here’s the first part of my review of some interesting machine learning (ML) papers I read in 2023.  As with the previous editions , this shouldn’t be considered a comprehensive review.  The papers covered here reflect my research interests and biases, and I’ve certainly overlooked areas that others consider vital.  This post is pretty long, so I've split it into three parts, with parts II and III to be posted in the next couple of weeks.    I. Docking, protein structure prediction, and benchmarking II. Large Language Models, active learning, federated learning, generative models, and explainable AI III. Review articles 2023 was a bit of a mixed bag for AI in drug discovery.  Several groups reported that the deep learning methods for protein-ligand docking weren’t quite what they were initially cracked up to be.  AlphaFold2 became pervasive, and people started to investigate, with mixed success, the utility of predicted protein structures.  There were reports of significant advanc
Read more

Generative Molecular Design Isn't As Easy As People Make It Look

Image
I was taken aback by a recent CNBC article entitled “ Generative AI will be designing new drugs all on its own in the near future ”.  I should know better than to pay attention to AI articles in the popular press, but I feel that even scientists working in drug discovery may have a skewed perception of what generative AI can and can’t do.  To understand exactly what’s involved, it might be instructive to walk through a typical generative molecular design workflow and point out a few things.  First, these programs are far from autonomous.  Even when presented with a well-defined problem, generative algorithms produce a tremendous amount of nonsense.  Second, domain expertise is essential when sifting through the molecules produced by a generative algorithm.  Without a significant medicinal chemistry background, one can’t make sense of the results.  Third, while a few nuggets exist in the generative modeling output, a lot of work and good old-fashioned cheminformatics are required to ext
Read more

AI in Drug Discovery - A Highly Opinionated Literature Review (Part II)

Image
Picking up where we left off in Part I , this post covers several other ML in drug discovery topics that interested me in 2023.  Some areas, like large language models, are new, and most of the work is at the proof-of-concept stage.  Others, like active learning, are more mature, and several groups are starting to explore nuances of the methods.   Here’s the structure of Part II.  4. Large Language Models 5. Active Learning 6. Federated Learning 7. Generative Models 8. Explainable AI 9. Other Stuff 4. Large Language Models The emergence of GPT-4 and ChatGPT brought considerable attention to large language models (LLMs) in 2023.  In November and December, several large pharmas held “AI Day” presentations featuring LLM applications for clinical trial data analysis. Many of these groups demonstrated the ability of LLMs to ingest large bodies of unstructured clinical data and subsequently generate tables and reports based on natural language queries.  Aside from some very brief demos on co
Read more