Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2014. / Cataloged from PDF version of thesis. / Includes bibliographical references. / Living cells integrate a vast array of protein-protein interactions (PPIs) to govern cellular functions. For instance, PPIs are critical to biosynthesis, nanostructural assembly, and in processing environmental stimuli through cell-signaling pathways. As fields such as synthetic biology and protein engineering mature they seek to mimic and expand the functions found in living systems that integrate PPIs. A critical feature to many PPIs that are integrated together to perform a complex function is orthogonality, i.e. PPIs that do not cross interact with each other. The engineering of orthogonal PPIs is thus an alluring problem. Since it not only tests our understanding of molecular specificity by having to stabilize and destabilize interactions simultaneously. The results of the design process can also have interesting applications in synthetic biology or bionanotechnology. The coiled coil, a rope-like structure made of helices, is a PPI ubiquitously found in biological systems and is an attractive fold for engineering orthogonal PPIs. Though the coiled coil is well studied, destabilization of undesired interactions still remains challenging. In this thesis I will discuss strategies for obtaining orthogonal PPIs, and describe the current sequence-to-structure relationships known about coiled coils. I will then introduce the computational multistate design framework, CLASSY, and explain how I applied it to the computational design of six orthogonal antiparallel homodimeric coiled coils. Five of these designed sequences were experimentally tested, of which only three of the sequences adopted the target antiparallel homodimer topology. All three of these sequences, as well as a previously designed antiparallel homodimer, were tested for cross reactivity in a pairwise manner. None of these sequences appeared to cross react. The sequences that failed to adopt the antiparallel topology highlight the need for improving our computational design framework. In the final chapter I will discuss strategies to improve our models, and applications for orthogonal antiparallel coiled coils. / by Christopher Negron. / Ph. D.

Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2016. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from student-submitted PDF version of thesis. / Includes bibliographical references (pages 139-156). / Microbes are tiny metabolic engines with large-scale effects on industry, the environment, and human health. Understanding how the micron-scale actions (and interactions) of individual microbes give rise to macro-scale consequences remains a major challenge in microbial ecology. However, for the most part, studies employ coarsegrained sampling schemes, which average over the heterogeneous microscopic structure of microbial communities. This has limited our ability to establish mechanistic links between dynamics occurring across these disparate spatial scales. However, such links are critical for (a) making sense of the tremendous extant microbial diversity on Earth, and (b) predicting how perturbations (e.g., global climate change) may influence microbial diversity and function. In this thesis, I characterize the structure and dynamics of wild bacterial populations in the ocean at spatial scales of tens of microns. I then employ a simple, two-strain laboratory model system to link (cooperative) inter-species interactions at local scales to emergent properties at larger scales, focusing on spatially connected meta-communities undergoing range expansions into new territory. This work encompasses diverse environments (ranging from well-mixed communities in the laboratory to individual crustaceans) and approaches (including mathematical modeling, highthroughput sequencing, and traditional microbiological experiments). Altogether, we find that the microscale environment inhabited by a microbe - that is, "what the neighborhood is like" and "who lives next to whom" - shapes the structure and dynamics of wild microbial populations at local scales. Moreover, these local interactions can drive patterns of biodiversity and function, even at spatial scales much larger than the length of an individual cell. Thus, our work represents a small step toward developing mechanistic theories for how microbes shape our planet's ecosystems. / by Manoshi Sen Datta. / Ph. D.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Computational and Systems Biology Program, 2013. / Cataloged from PDF version of thesis. "September 2013." Page 126 blank. / Includes bibliographical references. / The cholinergic innervation of the neocortex by afferent fibers originating in the nucleus basalis (NB) of the basal forebrain is implicated in modulating diverse neocortical functions including information processing, cortical plasticity, arousal and attention. To understand the physiological basis of these brain functions during cholinergic modulation, it is critical to identify the cortical circuit elements involved and define how their interactions contribute to cortical network dynamics. In this thesis, I present evidence showing how specific neuronal and glial cell types can be differentially modulated by acetylcholine (Ach), resulting in dynamic functional interactions during ACh-modulated information processing and cortical plasticity. Chapter 2 identifies somatostatin-expressing neurons as a dominant player in driving decorrelation and information processing through its intimate interactions with parvalbumin-expressing and pyramidal neurons. Chapter 3 highlights astrocytes and their interactions with pyramidal neurons as important drives for stimulus-specific cortical plasticity during cholinergic modulation. This is one of the earliest works that has mapped the functional role of distinct cell-types and their interactions to specific brain functions modulated by ACh, thereby setting the foundation for future studies to manipulate these specific functional interactions in both normal and diseased brains. / by Naiyan Chen. / Ph.D.

Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2018. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 153-164). / Protein-protein interactions (PPIs) play a major role in cellular function, mediating signal processing and regulating enzymatic activity. Understanding how proteins interact is essential for predicting new binding partners and engineering new functions. Mutational analysis is one way to study the determinants of protein interaction. Traditionally, the biophysical study of protein interactions has been limited by the number of mutants that could be made and analyzed, but advances in high-throughput sequencing have enabled rapid assessment of thousands of variants. The Keating lab has developed an experimental protocol that can rank peptides based on their binding affinity for a designated receptor. This technique, called SORTCERY, takes advantage of cell sorting and deep-sequencing technologies to provide more binding data at a higher resolution than has previously been achievable. New computational methods are needed to process and analyze the high-throughput datasets. In this thesis, I show how experimental data from SORTCERY experiments can be processed, modeled, and used to design novel peptides with select specificity characteristics. I describe the computational pipeline that I developed to curate the data and regression models that I constructed from the data to relate protein sequence to binding. I applied models trained on experimental data sets to study the peptide-binding specificity landscape of the Bc1-xL, Mc1-1, and Bf1-1 anti-apoptotic proteins, and I designed novel peptides that selectively bind tightly to only one of these receptors, or to a pre-specified combination of receptors. My thesis illustrates how data-driven models combined with high-throughput binding assays provide new opportunities for rational design. / by Vincent Xue. / Ph. D.

Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, February 2016. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 99-110). / The vast community of microbes that inhabit the human body, the human microbiota, is important to human health and disease. These microbes contribute to human metabolism, the development of the immune system and pathogen resistance, while imbalances among them have been associated with several diseases. In this work, I develop computational methods to gain key insights into the ecological principles that shape these communities. In the first chapter, I develop an evolutionary rate heuristic that leads to the discovery of a massive network of recently exchanged genes, connecting diverse bacteria throughout the human microbiota. Using this network, I examine the roles of phylogenetic distance, geographic proximity and ecological overlap in shaping rates of horizontal gene transfer. Of these factors, ecological similarity is the principal force shaping gene exchange. In the second chapter, I focus on the microbial communities within a person, identifying the factors that affect the stability of the human microbiota. Alpha-diversity is strongly correlated with stability, but the direction of this correlation changes depending on the body site or subject being examined. In contrast, beta-diversity is consistently negatively correlated to stability. I show that a simple equilibrium model explains these results and accurately predicts the correlation between diversity and stability in every body site, thus reconciling these seemingly contradictory relationships into a single model. In the final chapter, I explore the use of fecal microbiota transplantation (FMT) to treat recurrent Clostridium difficile infection. I develop a new method to infer the genotypes and frequencies of bacterial strains in metagenomics samples. I apply this method to a dataset covering twenty patients before and after FMT, uncovering key ecological rules that govern the colonization and growth of bacteria in human subjects after FMT. / by Christopher Scott Smillie. / Ph. D.

No description available.

Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2019 / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 210-226). / DNA methylation is an epigenetic mark that is highly conserved and important in diverse cellular processes, ranging from transposon silencing to genomic imprinting. In plants, DNA methylation is both mitotically and meiotically heritable, and changes in DNA methylation can be generationally stable and have long-lasting consequences. This thesis aims to improve understanding of DNA methylation dynamics in plants, particularly across generations and during reproduction. In the first project, I present an analysis of the generational dynamics of gene body methylation using recombinant inbred lines derived from differentially methylated parents. I show that while gene body methylation is highly generationally stable, changes in methylation state occur nonrandomly and are enriched in regions of intermediate methylation. / Important DNA methylation changes also occur during seed development in flowering plants, and these changes underlie genomic imprinting, the phenomenon of parent-of-origin specific gene expression. In plants, imprinting occurs in the endosperm, a seed tissue that functions analogously to the mammalian placenta. Imprinted expression is linked to DNA methylation patterns that serve to differentiate the maternally- and paternally-inherited alleles, but the mechanisms used to achieve imprinted expression are often unknown. I next explore imprinted expression and DNA methylation in Arabidopsis lyrata, a close relative of the model plant Arabidopsis thaliana. I find that the majority of imprinted genes in A. lyrata endosperm are also imprinted in A. thaliana, suggesting that imprinted expression is generally conserved. Surprisingly, a subset of A. lyrata imprinted genes are associated with a novel DNA methylation pattern and may be regulated by a different mechanism than their A. / thaliana counterparts. I then explore the genetics of paternal suppression of the seed abortion phenotype caused by mutation of a maternally expressed imprinted gene. Finally, I present the first large single-nuclei RNA-seq dataset generated in plants, reporting data from 1,093 individual nuclei obtained from developing seeds. I find evidence of previously uncharacterized cell states in endosperm, and examine imprinted expression at the single-cell level. Together, these projects contribute to our understanding of DNA methylation and imprinting dynamics during plant development, and highlight the strong generational stability of certain DNA methylation patterns. / by Colette Lafontaine Picard. / Ph. D. / Ph.D. Massachusetts Institute of Technology, Computational and Systems Biology Program

No description available.

<p>Vi har identifierat flexibelt utbyte och lagring av data i databaser, tillsammans med långvarig satsning på olika existerande och framtida modeller som nyckar till förståelse av det regler nätverk som utgör bron mellan geno- och fenotyp. Denna pilot studie av modellering av stora cellulära kontroll nätverk utgår från en intressant medicinsk frågeställning inom molekylär cellbiologi: Är framtvingad expression av Cdc6, aktivering av Cdk4/6 och Cdk2 tillräcklig för förankringsfri entré av cell cykelns S fas? Vi försöker konstruera en modell för att besvara denna fråga, på så sätt att vi kan detektera problem vid modellering av stora kontroll nätverk, diskutera implikationer och möjliga lösningar.</p><p>Vår modell är baserad på 1447 reaktioner och innehåller 1343 olika molekyler. Vi använde graf teori för att studera dess topologi och gjorde följande fynd: Nätverket är skalfritt och avtar enligt en potensfunktion, som var väntat baserat på tidigare arbeten. Nätverket består av ett stort väl förenat kluster. Det kan inte bli modulariserat i form av starka komponenter eller block i en användbar form. Detta eftersom vi fann en stor komponent eller ett stort block som innehöll majoriteten av alla molekyler och mer än hundra små komponenter eller block med en eller några molekyler. Vårt nätverk stämmer inte överens med en hierarkisk nätverks modell bestående av block förenade av cut-vertices.</p> / <p>We have identified flexible exchange and storage of data in databases, together with prolonged investment in different existing and future modelling formalisms as key issues in successful understanding of the regulatory network responsible for the connection between geno- and phenotype. This pilot study of modelling of large-scale regulatory networks starts with a medically interesting question from molecular cell biology: Is enforced expression of Cdc6, activation of Cdk4/6 and Cdk2 sufficient for anchorage-independent entry of the S phase of the cell cycle? We try to construct a model for answering this question, in such a way that we can reveal obstacles of large-scale regulatory modelling, discuss their implications and possible solutions.</p><p>Our model is based on 1447 reactions and contains 1343 different molecules. We used graph theory to study its topology and made the following findings: The network is scale-free and decays as a power-law, as could be expected based on earlier works. The network consists of one huge well connected cluster. It cannot be modularised into strong components or blocks in a useful way, since we get one big component or block containing a majority of all molecules and more than a hundred tiny components or blocks with one or a few molecules. Our network does not agree with a hierarchical network model consisting of blocks linked by cut-vertices.</p>

The completion of the human genome project has led to a flood of new genetic data, that has proved surprisingly hard to interpret. Network "guilt by association" (GBA) is a proven approach for identifying novel disease genes based on the observation that similar mutational phenotypes arise from functionally related genes. However, GBA has been shown to work poorly in genome-wide association studies (GWAS), where many genes are somewhat implicated, but few are known with very high certainty. In the first part of this work, I resolve this by explicitly modeling the uncertainty of the associations and incorporating the uncertainty for the seed set into the GBA framework. I demonstrate a significant boost in the power to detect validated candidate genes for Crohn’s disease and type 2 diabetes by comparing the predictions from my method to results from follow-up meta-analyses, with incorporation of the network serving to highlight the JAK--STAT pathway and associated adaptors GRB2/SHC1 in Crohn’s disease and BACH2 in type 2 diabetes. Consideration of the network during GWAS thus conveys some of the benefits of enrolling more participants in the GWAS study. More generally, we demonstrate that a functional network of human genes provides a valuable statistical framework for prioritizing candidate disease genes in GWAS-based studies. Furthermore, functional gene networks are not the only kind of information that can be used to predict gene--phenotype associations. In the second part of this thesis, I show that gene-phenotype associations in model species from species as distantly related to humans as E. coli is another valuable source of information, that can be mined using methods similar to those used in recommender systems. Finally, in the last part of this thesis, I present a machine learning formalism that combines the functional gene network and model species phenotype information. I show that this approach outperforms the state of the art methods for gene-phenotype association prediction using cross-validation. / text