The Bioinformatics and Beyond Podcast

Dr. Justin Siegel begins this episode by explaining what enzymes are, how they have evolved, and why Dr. Siegel is motivated to try to engineer enzymes to perform functions tailored to help humanity instead of to perform functions based on how they evolved in nature. He explains the primary goal of the work discussed and relating enzyme sequence to function. Dr. Siegel also explains how his work was the first of its kind by scaling up enzyme design to hundreds of mutants instead of dozens. 

We then dig into the details of Dr. Siegel’s work. We learn details of his study such as why his team chose to study the particular enzyme that was used to create a massive set of enzyme mutants. We hear the previous difficulty of doing a study like this on only one enzyme and what enabled this increase in the scale of enzyme design. We also hear about how the use of cloud labs was introduced into the project and why. 

Next, we hear all about the cloud lab aspect of this project. Dr. Siegel explains which parts of the enzyme mutant creation process were most challenging and benefited most to be moved to cloud labs. 

Finally, we learn about how machine learning was then applied to the large set of generated enzyme mutants. Dr. Siegel explains how the generated data allowed his team to test previous hypotheses about mutant enzymes and to start trying to predict the functions of enzymes from sequence. Dr. Siegel also comments on a finding of the paper that for highly conserved residues, if you change them, you lose the function. 

Learn more about Dr. Siegel’s work by reading the corresponding publication which you can find here: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0147596

Dr. Justin Siegel explains the past, present, and future of wet lab work and wet lab automation. We start by hearing a description of what it is like to work in a wet lab, covering the contrast between the excitement of seeing life changing results and the countless hours of monotony that is often involved to produce these results. 

We then begin discussing where automation will fit in to help alleviate the burden of long term monotonous work in the wet lab. We learn about the challenges of implementing automation in a lab, and hear about the dream that exists from the promise of automation versus the reality of implementing automation in an actual academic lab or in industry. We also hear Dr. Siegel’s take on the current state of implementing automation in an actual lab right now. We hear about the intricacies of implementing automation, such as discussing the pros and cons of different types of brands of robots, hearing about how lab robots can end up sitting unutilized or underutilized in academic labs, and considering practical questions that are involved when implementing automation. We end our discussion of robots that could be purchased with a discussion on Opentrons.  

Finally, we discuss cloud labs. Dr. Siegel starts by explaining what cloud labs are. Then, we hear about how a scientist would actually go about utilizing a cloud lab service. Dr. Siegel shares his thoughts on the potential promise of cloud labs and gives justification for the excitement surrounding this new approach. Dr. Siegel also shares his personal experience using cloud labs and how things like the accuracy and reliability of cloud labs can already make it a viable option for automating academic lab tasks. He also explains an unintended benefit of using cloud labs in that it allows researchers to spend more time thinking critically about the tasks that need to be done and how they will be done.

Dr. Afshin Beheshti begins this episode by explaining what microRNAs are and why they are emerging as an important area of biological research. He then explains how microRNAs relate to viruses, which is a recently developing area of research in this already young field of study. Dr. Beheshti then tells the story of how he started to discover that microRNAs could be a driver of COVID-19 infections. 

His story begins by using microRNA analysis tools to analyze COVID-19 infected patients from China which predicted a handful of microRNAs that could be involved in COVID-19 infection. He discusses how his team decided to focus on microRNA 2392 and how he continued to dig further into how it could be connected to COVID-19. His story then weaves through tales of successful collaborations with a large team of scientists that led to studying RNA samples from deceased COVID-19 patients, testing expressing the microRNA in healthy cells, analyzing multiple organs in COVID-19 infections, and testing a delivery system for a microRNA antagonist as a potential novel therapeutic.

We conclude with a quick discussion of the connection between microRNAs and space biology and space omics research.

Learn more about this work by checking out Dr. Beheshti's preprint on these topics: https://www.biorxiv.org/content/10.1101/2021.04.23.441024v4.abstract

This episode concludes the podcast’s series of episodes focused on space biology and space omics. NASA scientist Dr. Afshin Beheshti discusses the many high level hazards and corresponding molecular features of spaceflight throughout this episode. For instance, we begin with a discussion of the hazards of radiation and microgravity. Dr. Beheshti spends time explaining a high level view of what each hazard is, why it is a concern for spaceflight, and educates us on many useful and interesting pieces of information for each hazard. Further hazards discussed include confinement and isolation, hostile and closed environment, and distance from Earth.

After learning about all the high level hazards of extended living in space, we learn about how these hazards cause issues to human health through a series of lower level biological features. Dr. Beheshti again explains what these fundamental molecular features are, what techniques we have to study them, and ways we could overcome these problematic processes. These problematic molecular features include oxidative stress, DNA damage, mitochondrial damage, epigenetic and gene regulation changes, telomere-length dynamics, and microbiome shifts. 

We end by discussing how we can simulate and study the negative effects of space here on Earth and the future of spaceflight biology research. Dr. Beheshti explain how studies like "bed studies" and mountain climber studies can help simulate impacts on human health in space. Finally, I ask Dr. Beheshti for his view of the future. He explains NASA's surveys that can guide research and how omics research was identified as a future focus. We conclude with a discussion on the plan for Mars exploration and habitation.

For additional reading on this topic, check out Dr. Beheshti's recent Cell review: https://www.sciencedirect.com/science/article/pii/S0092867420314574

This episode continues our series of episodes on space biology and #SpaceOmics with Dr. Tejaswini Mishra. Dr. Mishra introduces The NASA Twins Study, a cornerstone scientific work where two twin astronauts were monitored, with one twin traveling to space, and one staying on earth. Dr. Mishra explains the importance of studying long term spaceflight missions, how The NASA Twins Study was set up in a particularly great way to study spaceflight impact, the many different types of data collected and analyzed, and some of the results found by the study.

During the episode, Dr. Mishra explains many of the types of data collected such as microbiome and telomere data. After covering the types of data, we explore some of the main results such as the first ever test of a vaccination in space. Dr. Mishra then explains more in depth on changes seen during spaceflight such as telomere length, gene expression, DNA damage, cognitive function, and more. We discuss how concerning the various changes that occur in space could be for astronauts, such as becoming hypoxic. Dr. Mishra also explains pointers to the kinds of things we should focus on when we go deeper into space for understanding the impact on the human body. Finally, we summarize the main messages of the paper and hear Dr. Mishra’s thoughts on the future of space research. 

The NASA Twins Study can be found at: https://science.sciencemag.org/content/364/6436/eaau8650

For people who work in the life sciences, a very common occurrence is for folks who work on the "wet" side of research, largely doing bench work, to become interested in or start wanting to transition to doing more "dry" research, like computational research in bioinformatics. In this special episode, dedicated to those thinking about transitioning from "wet" lab work to doing more "dry" lab type work, my guest Dr. Willian da Silveira explains his own transition from a full bench scientist to a full time bioinformatician. Dr. da Silveira also answers many questions from the bioinformatics subreddit on this topic. Following Dr. da Silveira's explanation of his career trajectory and his own shift from "wet" lab work to "dry" lab work, I ask a series of questions from the bioinformatics subreddit seen below, with time stamps included:

[19:00] Bioinformatics subreddit questions begin.

[20:00] What general stats and technical requisites are necessary to transition from wet lab to dry lab work?

[23:30] Is it boring to only do data analysis versus conducting lab experiments?

[27:50] Should you transition early, for example during a masters or PhD program, or can it be done later?

[33:30] Does the transition need to be forced or does it happen more often by chance?

[35:10] Is there a downside to being self-taught as a bioinformatician?

[36:20] What are the upsides of picking up bioinformatics later on, starting as a wet lab scientist first?

[40:25] How to get accepted into a bioinformatics PhD program with no formal CS education?

[42:07] What about dry lab to wet lab transitioning?

[48:07] How do you get your foot in the door when switching from the wet lab to a dry lab with little or no dry lab experience on your CV?

[50:45] If you do feel stuck, would the best route be to go ahead and pick up some formal education like a paid masters degree?

[52:48] Would it make sense to transition to dry lab work given employment and financial considerations?

Finally, to end the discussion, I ask Willian what he thinks the ideal mix of wet lab and dry lab experience might look like.

In this episode we begin discussing the biology of spaceflight with Dr. Willian da Silveira. We start by hearing the story of how Dr. da Silveira's recent high profile space omics paper (https://www.cell.com/cell/pdf/S0092-8674(20)31461-6.pdf) came to be. He first describes the NASA GeneLab and how he got involved, and how his story of this paper began with an analysis of some liver transcriptomics data. We hear about all the different types of data used in this study, including epigenetics and metabolomics data. Dr. da Silveira discusses how to try to incorporate and work with this many types of data all at the same time. He then further elaborates and explains data like epigenetics and metabolomics. 

After discussing all the different types of data, and how to try to analyze all the data together, Dr. da Silveira talks more about the biological side of some of the data, for instance discussing rodent data and human cell lines. Finally, we discuss the results of his paper and how all the data analysis point to a central hub of the impact of spaceflight, with mitochondrial stress acting as this central hub. We conclude with a discussion of the principal risks to humans when they go to space and what Dr. da Silveira sees coming for the future of space omics research.

Link to Dr. da Silveira's recent publication: https://www.cell.com/cell/pdf/S0092-8674(20)31461-6.pdf

Link to spaceflight impact review paper mentioned: https://pubmed.ncbi.nlm.nih.gov/33242416/

Dr. Hayden Metsky begins by introducing the ADAPT method for doing large-scale detection of viruses. ADAPT is a computational method that aids the design of CRISPR-based viral testing. He then discusses the motivation for ADAPT and how it relates to his previous works like CATCH. In comparing ADAPT to other work, Dr. Metsky discusses, for instance, differences between CRISPR-based testing and more traditional testing like qPCR. In discussing the challenges of designing diagnostic tests and detection assays, Dr. Metsky then describes how he breaks these challenges into three different components. 

Dr. Metsky goes on to talk about how they designed an assay based around the Cas13 enzyme. He describes how they used this approach for targeting viruses and explains how they designed a large-scale library of 20,000 pairs of target viral sequences and guide RNAs. He then explains how they used this library as training data for a machine learning model. He also explains his thought process of designing and training the convolutional neural network model that they ended up using for predicting how well the guide RNAs would work. 

As our conversation continues, Dr. Metsky points out an interesting observation that his team made while working on this project. He points out that it could be the case that it may not be best to only design diagnostics around a highly or universally conserved region. He explains that taking into account other considerations, like how well the diagnostic technology works for a particular target sequence, may produce even better results. He also points out how it can be really challenging to only consider the highly conserved or totally conserved regions because those regions are going to be the most likely to be shared by other viruses or organisms which induce false positives in tests. Dr. Metsky explains his thought process for how you take a problem like this, figure out the characteristics of the problem, and match it well to a closely related problem or other scientific works, explaining the process of figuring out how to optimize the final objective function in ADAPT. Final topics include a discussion on the speed of ADAPT and the availability of the software. 

To learn more about ADAPT, you can read the ADAPT manuscript at https://www.biorxiv.org/content/10.1101/2020.11.28.401877v2 or visit the software page at https://github.com/broadinstitute/adapt

Dr. Hayden Metsky begins the episode by describing his goal of being able to harness sequenced viral genomes to computationally design diagnostics, therapies, and vaccines. He discusses the value of having methods available that can handle all available genomic data for diverse species for diagnostics and therapies. Next, we learn how CRISPR can be used in a diagnostics setting. Dr. Metsky explains how collateral cleavage broadens the use of CRISPR beyond simply being a tool for genome editing. Advantages and disadvantages of CRISPR-based diagnostics techniques are discussed versus, for example, a more traditional qPCR approach. The discussion then moves on to the computational component of the diagnostics design problem. Dr. Metsky discusses his 2019 Nature Biotechnology paper on the CATCH method for use in hybridization capture and his progression of work in this area (see https://www.nature.com/articles/s41587-018-0006-x). Finally, we discuss his work in designing diagnostics for SARS-CoV-2, CRISPR-based tests being able to gain widespread adoption, and expanding this work beyond viruses to include bacteria as well.