Links
12/16/2022 update: CASP15 just occurred and incredible things are happening in the protein structure prediction/design ML field. This article sums up what in my opinion are some of the most exciting recent developments. Several git-hub accounts provide a comprehensive list of papers and demonstrations for the myriad of techniques that have sprung up (see below). Some specific highlights that I think have super high utility include Alpha-pulldown, Alpha-fill, Deep-learning language models to improve sequence alignment, and @sokrypton's End-to-end learning of multiple sequence alignments methods.
Several lists linking various deep learning codebases, papers, and information – start here
A huge list of papers on protein design using deep learning
Another list of awesome deep bio papers and applications
At the same time the realm of CryoEM has been steadily marching forward. Exciting new developments include @SjorsScheres' ModelAngelo, and generally crazy improvements in resolution.
Several lists linking various deep learning codebases, papers, and information – start here
A huge list of papers on protein design using deep learning
Another list of awesome deep bio papers and applications
At the same time the realm of CryoEM has been steadily marching forward. Exciting new developments include @SjorsScheres' ModelAngelo, and generally crazy improvements in resolution.
Obviously AlphaFold is pretty great. For the uninitiated, here is a link to the google collab notebook for running a round of protein prediction or the alternative Rosetta web server for protein folding. Here is another great set of notebooks by Dr. Ovchinnikov that are worth mentioning as well.
Here are some notebooks for running MD simulations using CHARMM and Amber to set up the simulation and initialized the force field and water model, and OpenMM as the simulation engine.
A series of notebooks to teach protein engineering using PyRosetta from the Grey lab.
While I am just an armchair immunologist (for now), in my opinion this is one of the most exciting thing going on in science today:
Dendritic cell vaccines have the potential to cure cancers, prevent malaria, HIV, TB, and countless other diseases that modern medicine currently struggles with. I think they are particularly exciting as compared to other strategies for immune modulation because they are incredibly efficient at presenting antigens and they are at the beginning of the complex cascade of cellular interactions that occur in a natural immune response which may activate cell types like CD8+ T-cells and promote T-cell memory. Hopefully this means reliable and tunable activation of immune responses that are appropriately tailored to diseases which currently evade our immune system.
One cool example of this strategy used radiation inactivated Plasmodium parisites (Malaria), in addition to an antibody that bound the endocytic DC receptor Clec9A. This antibody was decorated with an antigen which was known to provoke an immune response against the Plasmodium. By targeting this antigen to the Clec9A+ DCs they were able to promote the enhancement of immune memory against the Plasmodium.
Other aspects of immunology that interest me include antigen presentation in cancer, receptor-mediated endocytosis and antigen trafficking, immune synapses, RNA and DNA vaccines, nanobodies and intrabodies, and vaccine scaffold engineering.
Some of my favorite papers include:
More favorites:
I would be lying if I said I do not struggle with manifold based machine learning and the finer points of deep learning in general. I am slowly educating myself on the math behind it on youtube when I have time. Here is an approachable paper on the topic:
Another fascinating interdisciplinary topic I would love to dive into is the application of deep learning in biochemistry. Here are some examples:
A Deep Learning Approach to Antibiotic Discovery
Improved Protein Structure Prediction Using Potentials from Deep Learning
I am slowly educating myself, and I really am benefiting from some great online teachers and courses. I have taken Jose Portilla's course on udemy on Tensor Flow, however I felt that it did not go deep enough (hehe) so I have started Andrew Ng's course on Coursera, and found that the course load was too much for me at the moment, but I plan to get back to it as soon as the PhD is over. Then I would like to get my own local instance of AlphaFold running (whenever I can afford a workstation with 2TB of memory).
One of my favorite blogs:
Derek Lowe's excellent blog on medicinal chemistry
Matt Might has a great blog on grad school and academia in general:
Matt Might's blog post on PhDs
Matt Might's blog post on resolutions for grad students
An article that every scientist should read:
The importance of stupidity in scientific research
A great place to learn about the structure of proteins: The protein data bank
A great pop sci article on protein engineering and protein folding
A great post on why too many meetings can kill productivity
I hear it said a lot these days that education needs reforming. I disagree. I think mainstream education and our methods of certification need reforming...
Great educational youtube channels:
A few articles on a topic that can never get too much air time:
Nature article: Time to talk about why so many postgrads have poor mental health
Nature biotechnology paper: Evidence for a mental health crisis in graduate education
Science article: Ph.D. students face significant mental health challenges
We are all being tracked!
Here are some notebooks for running MD simulations using CHARMM and Amber to set up the simulation and initialized the force field and water model, and OpenMM as the simulation engine.
A series of notebooks to teach protein engineering using PyRosetta from the Grey lab.
While I am just an armchair immunologist (for now), in my opinion this is one of the most exciting thing going on in science today:
Dendritic cell vaccines have the potential to cure cancers, prevent malaria, HIV, TB, and countless other diseases that modern medicine currently struggles with. I think they are particularly exciting as compared to other strategies for immune modulation because they are incredibly efficient at presenting antigens and they are at the beginning of the complex cascade of cellular interactions that occur in a natural immune response which may activate cell types like CD8+ T-cells and promote T-cell memory. Hopefully this means reliable and tunable activation of immune responses that are appropriately tailored to diseases which currently evade our immune system.
One cool example of this strategy used radiation inactivated Plasmodium parisites (Malaria), in addition to an antibody that bound the endocytic DC receptor Clec9A. This antibody was decorated with an antigen which was known to provoke an immune response against the Plasmodium. By targeting this antigen to the Clec9A+ DCs they were able to promote the enhancement of immune memory against the Plasmodium.
Other aspects of immunology that interest me include antigen presentation in cancer, receptor-mediated endocytosis and antigen trafficking, immune synapses, RNA and DNA vaccines, nanobodies and intrabodies, and vaccine scaffold engineering.
Some of my favorite papers include:
- Electron-Microscopy-Based Epitope Mapping Defines Specificities of Polyclonal Antibodies Elicited during HIV-1 BG505 Envelope Trimer
- Structure of the Ebola virus glycoprotein bound to an antibody from a human survivor
- ImmunizationComputation-Guided Backbone Grafting of a Discontinuous Motif onto a Protein Scaffold
- Proof of principle for epitope-focused vaccine design
- ArticleLiver-Resident Memory CD8+T Cells Form a Front-Line Defense against Malaria Liver-Stage Infection
- Towards superior dendritic-cell vaccines for cancer therapy
- Multi-particle cryo-EM refinement with M visualizes ribosome-antibiotic complex at 3.5 Å in cells
- Single-particle cryo-EM at atomic resolution
More favorites:
- Cell-permeable nanobodies for targeted immunolabelling and antigen manipulation in living cells
- Just passing through
- Generation of synthetic nanobodies against delicate proteins
I would be lying if I said I do not struggle with manifold based machine learning and the finer points of deep learning in general. I am slowly educating myself on the math behind it on youtube when I have time. Here is an approachable paper on the topic:
Another fascinating interdisciplinary topic I would love to dive into is the application of deep learning in biochemistry. Here are some examples:
A Deep Learning Approach to Antibiotic Discovery
Improved Protein Structure Prediction Using Potentials from Deep Learning
I am slowly educating myself, and I really am benefiting from some great online teachers and courses. I have taken Jose Portilla's course on udemy on Tensor Flow, however I felt that it did not go deep enough (hehe) so I have started Andrew Ng's course on Coursera, and found that the course load was too much for me at the moment, but I plan to get back to it as soon as the PhD is over. Then I would like to get my own local instance of AlphaFold running (whenever I can afford a workstation with 2TB of memory).
One of my favorite blogs:
Derek Lowe's excellent blog on medicinal chemistry
Matt Might has a great blog on grad school and academia in general:
Matt Might's blog post on PhDs
Matt Might's blog post on resolutions for grad students
An article that every scientist should read:
The importance of stupidity in scientific research
A great place to learn about the structure of proteins: The protein data bank
A great pop sci article on protein engineering and protein folding
A great post on why too many meetings can kill productivity
I hear it said a lot these days that education needs reforming. I disagree. I think mainstream education and our methods of certification need reforming...
Great educational youtube channels:
- A great course on Cryo Electron Microscopy led by Sjors Scheres with MRC labs. I learned a ton from this one.
- Getting started in CryoEM with professor Grant Jensen: another wildly helpful set of videos for learning CryoEM
- The Meiler lab Rosetta protein engineering tutorials are now available on YouTube
- 3blueonebrown is an amazing youtube channel for learning math!
- AK lectures is a great channel for learning biochemistry or organic chemistry and also has loads of other content
- Bob and Brad is an amazing channel for physical therapy and fitness
- NileRed is an awesome channel for watching very entertaining chemistry reactions
- The most comprehensive C++ tutorials I have found
- A crystallography course by MRC labs
- Some more videos by SBgrid Consortium on various software packages
- A great blog explaining Markov-chains
A few articles on a topic that can never get too much air time:
Nature article: Time to talk about why so many postgrads have poor mental health
Nature biotechnology paper: Evidence for a mental health crisis in graduate education
Science article: Ph.D. students face significant mental health challenges
We are all being tracked!