Tuesday, September 27, 2016

Lecture 10: Collin Tokheim (Rachel Karchin Lab)

Hotspots in Cancer.
Missense mutations are perhaps the most difficult mutation type to interpret in human cancers. Truncating loss-of-function mutations and structural rearrangements generate major changes in the protein product of a gene, but a single missense mutation yields only a small change in protein chemistry. The impact of missense mutation on protein function, cellular behavior, cancer etiology, and progression may be negligible or profound, for reasons that are not yet well understood. Missense mutations are frequent in most cancer types, accounting for approximately 85% of the somatic mutations observed in solid human tumors, and the cancer genomics community has prioritized the task of identifying important missense mutations discovered in sequencing studies. Whole exome sequencing (WES) studies of cancer have created new opportunities to better understand the importance of missense mutations. This enormous collection of data now allows detection of patterns with power that was unheard of a few years ago.

Comparison of hotspot detection in the TSG FBXW7 in 1D and 3D [1].
In their recent work, Collin Tokheim from Johns Hopkins Biomedical Engineering and his colleagues used The Cancer Genome Atlas mutation data and identified 3D clusters of cancer mutations ("hotspot regions") at amino-acid-residue resolution in 91 genes, of which 56 are known cancer-associated genes. The hotspot regions identified by their method are smaller than a protein domain or protein– protein interface and in many cases can be linked precisely with functional features such as binding sites, active sites, and sites of experimentally characterized mutations. The hotspot regions are shown to be biologically relevant to cancer, and they discovered that there are characteristic differences between regions in the two types of driver genes, oncogenes and tumor suppressor genes (TSG). These differences include region size, mutational diversity, evolutionary conservation, and amino acid residue physiochemistry. For the first time, the research team quantifies why the great majority of well-known hotspot regions occur in oncogenes. Because hotspot regions in TSGs are larger, more heterogeneous than those in oncogenes, they are more difficult to detect using protein sequence alone and are likely to be underreported. The results indicate that protein structure–based 3D mutation clustering increases power to find hotspot regions, particularly in TSGs.

Good reads on the subject:
[1] Tokheim, Collin, et al. "Exome-scale discovery of hotspot mutation regions in human cancer using 3D protein structure." Cancer research (2016): canres-3190.
[2] Perdigão, Nelson, et al. "Unexpected features of the dark proteome." Proceedings of the National Academy of Sciences 112.52 (2015): 15898-15903.
[3] Kamburov, Atanas, et al. "Comprehensive assessment of cancer missense mutation clustering in protein structures." Proceedings of the National Academy of Sciences 112.40 (2015): E5486-E5495.
Collin Tokheim got his bachelor's degree in biomedical engineering at the University of Iowa. After a short stint working in an RNA genomics lab, he came to Hopkins to work on a Ph.D. in Biomedical Engineering. He currently works in Rachel Karchin's lab doing computational research applied to cancer genomics. Collin's current research focuses on how protein structures used on an exome-scale can inform which missense mutations are likely drivers of cancer.

Tuesday, September 13, 2016

Lecture 9: Max C. Klein (Elijah Roberts Lab)

An epigenetic landscape
(from Epigenetics Unraveled).
Analyzing rare events in biology. 
Changes in cellular phenotype can often be tightly correlated with changes in biochemistry. The macroscopic (phenotypical) and the microscopic (biochemical) descriptions of these phenotype switching events can be unified through the theoretical framework of epigenetic landscapes (EL). In approximate terms, the EL of a cellular system is a map that describes every possible system state (expressed in terms of a count of relevant DNA, proteins, etc.) and the probability of the system being in each of these states. An EL can be used to completely describe the dynamics of its associated system, allowing for a deep level of understanding and, potentially, control of the system. It is possible to take the list of chemical species and reactions involved in a cellular process and, using theory and modeling, generate the corresponding EL. Unfortunately, the computer time required to generate an EL using the standard stochastic simulation method, called Brute Force Sampling (BFS), makes it difficult to generate the EL of even a relatively simple multi-state system.

The Forward Flux Sampling (FFS) method, which belongs to a larger family of Enhanced Sampling methods, was previously developed by ten Wolde and coworkers to address this computational limitation. Although FFS can speed up EL simulations by multiple orders of magnitude, there is a commensurate increase in the complexity of simulation setup. Specifically, there are a number of novel free parameters that the simulation user must specify. Each of these parameters has a significant and indirect influence on the precision of an FFS simulation’s final results. It is, therefore, desirable to develop a version of the FFS algorithm in which the relationship of parameter choice to error is made explicit.

Max Klein and Dr. Elijah Roberts from Johns Hopkins Biophysics have designed and implemented a supercomputer-compatible version of the Forward Flux Enhanced Sampling method for stochastic simulation of biochemical networks. Their version greatly simplifies parameter choice and simulation setup relative to the base Forward Flux algorithm. A user of the program needs only to specify the desired level of precision error. The program will determine a set of parameters to use that will achieve this target error level while optimizing the simulation run time. The Forward Flux implementation has been tested and verified in terms of its ability to calculate switching rate constants and epigenetic landscapes. A preview build of the code is currently available here.

Related Articles:
[1] Dickson, Alex, and Aaron R. Dinner. "Enhanced sampling of nonequilibrium steady states." Annual review of physical chemistry 61 (2010): 441-459.
[2] Allen, Rosalind J., Daan Frenkel, and Pieter Rein ten Wolde. "Forward flux sampling-type schemes for simulating rare events: Efficiency analysis." The Journal of chemical physics 124.19 (2006): 194111.
[3] Becker, Nils B., and Pieter Rein ten Wolde. "Rare switching events in non-stationary systems." The Journal of chemical physics 136.17 (2012): 174119.
[4] Gardner, Timothy S., Charles R. Cantor, and James J. Collins. "Construction of a genetic toggle switch in Escherichia coli." Nature 403.6767 (2000): 339-342.
Max C. Klein received his B.A. in Physics, from Reed College, Oregon in 2013. After graduation, he joined the Program in Molecular Biophysics at Johns Hopkins University. Currently, with Dr. Elijah Roberts from the Biophysics Department, Max is developing new methods to simulate decision making in cells using a hybrid multi-CPU/multi-GPU computational architecture.

Tuesday, September 6, 2016

Lecture 8: Dr. Yasser Aboelkassem (Natalia Trayanova Lab)

The heart of the matter. 
Mesh representation of the
ventricles of the heart
Ventricular relaxation occurs as intracellular calcium drops to resting levels. Under low calcium conditions, contraction is inhibited by the troponin/tropomyosin complex. However, experimental evidence has long suggested that some degree of actin-myosin interaction is possible even in the absence of calcium. Under calcium-free conditions, as many as 5% of actin binding sites are occupied by myosin, according to some estimates made from solution studies of purified myofilament components. Despite abundant in vitro evidence for calcium-independent activation (CIA), its relevance to in vivo cardiac function is not clear. Striated muscle preparations can produce small amounts of actin-myosin-based force under low calcium conditions, especially near physiological temperatures. This suggests that residual actin-myosin cross bridges resist diastolic filling, adding to the resistance provided by other structures such as collagen and titin. However, distinguishing the contributions of these various factors is technically challenging, and cross bridge-based diastolic stiffness remains controversial.

Schematic diagram of model components and states [1].
In their recent study, Dr. Yasser Aboelkassem and his colleagues investigate CIA using computational analysis by adding a structurally motivated representation of this phenomenon to an existing myofilament model, which allowed predictions of CIA-dependent muscle behavior. The researchers found that a certain amount of CIA was essential for the model to reproduce reported effects of nonfunctional troponin C on myofilament force generation. Consequently, those data enabled estimation of ΔGCIA, the energy barrier for activating a thin filament regulatory unit in the absence of calcium. Using this estimate of ΔGCIA as a point of reference (∼7 kJ mol−1), they examined its impact on various aspects of muscle function through additional simulations. CIA decreases the Hill coefficient of steady-state force while increasing myofilament calcium sensitivity. At the same time, CIA has minimal effect on the rate of force redevelopment after slack/restretch. Simulations of twitch tension show that the presence of CIA increases peak tension while profoundly delaying relaxation. We tested the model’s ability to represent perturbations to the calcium regulatory mechanism by analyzing twitch records measured in transgenic mice expressing a cardiac troponin I mutation (R145G). The effects of the mutation on twitch dynamics were fully reproduced by a single parameter change, namely lowering ΔGCIA by 2.3 kJ mol−1 relative to its wild-type value. The analyses of Dr. Aboelkassem's and his team suggest that CIA is present in cardiac muscle under normal conditions and that its modulation by gene mutations or other factors can alter both systolic and diastolic function.
[1] Aboelkassem, Yasser, et al. "Contributions of Ca 2+-Independent Thin Filament Activation to Cardiac Muscle Function." Biophysical journal 109.10 (2015): 2101-2112.
Dr. Yasser Aboelkassem obtained his BSc. in Aerospace Engineering from Cairo University, Egypt and received his master's degree in Mechanical Engineering from a joint program between Concordia-McGill University, Montreal, Canada. After graduation, Dr. Aboelkassem moved to Virginia Tech where he obtained a Master in Applied Mathematics degree and a Ph.D. in Engineering Science and Mechanics in 2012. He did his first postdoctoral training working in cardiac mechanics at the department of Biomedical Engineering at Yale University. Currently, he is a postdoc research associate with Natalia Trayanova at the Institute of Computational Medicine, Johns Hopkins University.

Dr. Aboelkassem has published 16 first-author papers and co-authored 6 papers in peer-reviewed international journals. His current research focus is the multiscale modelling of cardiac thin filament activation.