Asa Ben-Hur - US grants

Colorado State University is awarded a grant to develop machine learning methods for detecting alternative splicing in plants and to experimentally validate selected predictions. Alternative splicing has an important role in proteome diversity and gene regulation. Recent studies of large scale EST/cDNA datasets have revealed that the prevalence of alternative splicing in plants is much larger than expected, reaching around 30% of the genes, which is still significantly less than in human and mouse. This is primarily due to the much smaller amount of cDNA/EST data that is available in plants. Therefore we are likely far from the true extent of alternative splicing in plants. In human and mouse, several projects have made non-EST-based predictions of alternative splicing; none have been reported in plants to our knowledge. To fill this gap, the PIs will develop computational tools to predict novel alternative splicing events and the cis-elements involved in regulated alternative splicing. Alternative splicing in plants has different characteristics than in animals, and the proposed computational and experimental work will help elucidate the mechanistic basis for these differences. The initial focus will be in Arabidopsis, and the methods will be extended to rice and other plants for which genome and EST data are available. The end-results of the proposed research will be the creation of a web-accessible database of predicted and validated alternative splicing events and cis-elements; the software developed during the course of this project will be made available for researchers interested in predicting alternative splicing in other plant species.

Colorado State University is awarded a grant to develop machine learning methods for predicting protein function. The availability of protein function annotations supports the everyday work of biologists in multiple areas---from biomedical discovery to the study of plant drought resistance, and the design of bacteria useful in biofuel production. Assigning function to proteins in sequenced genomes is a major undertaking, and with new organisms being sequenced daily, experimentally determining the function of all the proteins in those organisms is not practical, requiring computational assignment of function to proteins that have not been studied in the lab. Computational scientists have been considering the problem of function prediction for over two decades. Yet, the basic methodology for protein function prediction has not changed much during this time and remains that of "annotation transfer" from proteins with a known function using a method for sequence comparison such as BLAST. Protein function prediction has several properties that make it difficult to apply state-of-the-art machine learning methods to this problem, such as the large number of potential functions (thousands of possible terms), the fact that proteins can have multiple functions, and the hierarchical relationship between terms in the Gene Ontology (GO), which is the standard system of keywords used to describe protein function. In this work the problem of annotating proteins with GO terms will be explicitly modeled as a hierarchical classification problem using the methodology of "kernel methods for structured outputs", which allows the modeling of complex prediction problems. This methodology will allow the PIs to integrate a variety of genomic information - sequence data, gene expression, protein-protein interactions, and information mined from the biological literature. The award will lead to a function prediction method with state-of-the-art accuracy. The project will have broad impact by providing the GOstruct method to the bioinformatics and biology communities in the form of downloadable software and an online-accessible function prediction server. Education will be impacted through the incorporation of the tool into new courses in programming for biologists and on kernel methods.

Biosensing using advanced semiconductor technologies and methods based on next generation DNA sequencing generate information about biological systems at an unprecedented scale. This National Science Foundation Research Traineeship (NRT) award prepares master's and doctoral students at Colorado State University with the tools to extract meaning from huge amounts of sensor and genomics data. Through innovative curricula and internships, trainees will learn to handle the new computational, statistical, mathematical, and engineering challenges that biologists, computer scientists, and engineers are unable to overcome alone. This traineeship program contributes to a new generation of scientists and engineers who are able to tackle complex problems related to large datasets in a variety of disciplines. Moreover, this program builds a community of university, K-12 schools, community colleges, and industry for a wider participation of effective scientific discovery, teaching, and learning.

Research activities integrated with the training program will address themes critical to advance understanding of some fundamental data-related questions facing biological sciences. Interdisciplinary teams will tackle research involving biological sensors, detection of microbes, regulation of gene expression, and evolutionary genomics and genome assembly. Faculty across engineering, life sciences, computer science, mathematics, and statistics will develop a flexible, customizable collection of training modules, to create a training experience personalized for each student regardless of their background. Trainees will develop the skills and tools to process, analyze, visualize, and understand large datasets from biosensing and next generation DNA sequencing. Additionally, the program incorporates a wide range of transferrable skills training so that trainees will be well equipped to engage and lead data-centric research within or outside academia.

Proteins perform many cellular functions, made possible by complex networks of interactions; knowing the location of the interaction sites on the proteins is key for understanding exactly how they work. Important applications include designing drugs and therapeutic agents. Experimental techniques for determining the interfaces between proteins are expensive and time consuming, so computational structural biologists seek to predict these mathematically. Current prediction methods use a limited number of features hand-crafted by an expert. An alternate approach is to learn the important features directly from all of the data, using a method called deep neural networks. This proposal explores a combined approach: use expert intuition for some features but add the power of unsupervised learning with deep neural networks to learn additional, novel features. The results will enrich the way protein structural features are understood in terms of their functional properties, whether those are catalytic sites, protein-binding sites or other sites important to the protein structure. Certainly in the prediction of protein structure itself machine learning scoring methods are showing great promise. Aspects of the research will be used in courses offered through a recently awarded NSF-NRT training grant, The training grant establishes an interdisciplinary program at the interfaces of biology, engineering, math/statistics and computer science. The program prepares students for a variety of career paths. Research and education experiences will provide students with valuable expertise in a computational area that is highly valued by top technology firms, such as Google and Facebook, which have research teams exploring the possibilities of deep neural networks.

This work proposes a paradigm shift in the field of protein interface prediction and scoring: from hand-crafted features and standard off-the-shelf classifiers to an approach that augments existing features with automatic learning of the features that characterize the 3-d structures of proteins, combined with the use of learning algorithms that are specifically designed for the characteristics of the problem. The proposed approach has multiple novel aspects: the proposed learning approach leverages information contained in the entire protein data bank (PDB) to learn features that characterize protein structures at multiple scales and levels of abstraction. It introduces a novel neural network architecture and regularization terms that constrain the solution towards biologically relevant results. The primary alternative to this machine learning-based interface prediction uses docking simulations; however, current docking energy functions are not accurate enough, so that a near-native solution is often not ranked high enough on the list of outcomes to be useful. Extensions of the proposed architectures for interface prediction will be employed for re-scoring docking solutions to improve their predictive success. A workflow that integrates docking and machine learning-based interface prediction and scoring is proposed to explore the synergism between these tasks. Information on the progress made on the project is available through the project website: http://www.cs.colostate.edu/~asa/projects.html.

Project Summary/Abstract The Quantitative Cell and Molecular Biology (qCMB) Training Program will provide comprehensive and flexible training in quantitative approaches in order to facilitate transdisciplinary and collaborative research in the broad field of Cell & Molecular Biology. The Program will produce outstanding graduates with the skills to meet the computational and professional demands of modern life science research. The objectives include to produce more graduates able to apply computational approaches to solve biological problems, build and sustain an accessible curriculum to provide training in this area, encourage and support transdisciplinary collaborations, provide training and professional development opportunities to better prepare students for the workforce, and facilitate completion by fostering a supportive and inclusive environment. The qCMB program is designed to support students engaging in collaborative research projects requiring the generation and analysis of large biological datasets particularly those generated through sequencing, imaging and flow cytometry. It combines recently developed computational and quantitative coursework with a new gateway course developed for this program, and innovative approaches to encourage reproducible research. Progress will be monitored through annual reports, seminars and committee meetings. Trainees will be prepared for entry to the workforce through personalized career mentoring, along with professional skills training, mentoring from industry professionals and guided access to internship experiences. Trainees will be welcomed and supported regardless of race, gender, age, sexual orientation or disability. The program will facilitate this through peer mentoring, established events such as an Annual Picnic, and Poster Symposium. We will build a qCMB community through monthly meetings, and a summer symposium/retreat. Preceptors will be supported through required training that will hone their mentoring skills and convey strategies to ensure mentees engage in reproducible research. Existing and new collaborations between preceptors working in areas such as Biology of Single Cells, Post-Transcriptional Control, Genome Architecture & Function, and Protein Structure & Function will provide students with opportunities to engage in transdisciplinary quantitative/computational research that requires comprehension of cell and molecular biology. Preceptor mentoring plans and effectiveness will be evaluated by the qCMB leadership. Potential trainees (8-10 each year) are recruited into the existing interdisciplinary Cell & Molecular Biology PhD Program, and rotate during their first year to identify a lab/preceptor that fits their research interests and mentoring needs. At the end of the first year, three trainees entering qCMB labs will be selected for support through this award and each will be supported for two years before transitioning to the preceptor's research grant or other fellowship for stipend support. The award will support a total of 15 trainees, who will transition to positions in academia, or industry that require data analysis, collaboration and research experience.

Project Summary/Abstract The Quantitative Cell and Molecular Biology (qCMB) Training Program will provide comprehensive and flexible training in quantitative approaches in order to facilitate transdisciplinary and collaborative research in the broad field of Cell & Molecular Biology. The Program will produce outstanding graduates with the skills to meet the computational and professional demands of modern life science research. The objectives include to produce more graduates able to apply computational approaches to solve biological problems, build and sustain an accessible curriculum to provide training in this area, encourage and support transdisciplinary collaborations, provide training and professional development opportunities to better prepare students for the workforce, and facilitate completion by fostering a supportive and inclusive environment. The qCMB program is designed to support students engaging in collaborative research projects requiring the generation and analysis of large biological datasets particularly those generated through sequencing, imaging and flow cytometry. It combines recently developed computational and quantitative coursework with a new gateway course developed for this program, and innovative approaches to encourage reproducible research. Progress will be monitored through annual reports, seminars and committee meetings. Trainees will be prepared for entry to the workforce through personalized career mentoring, along with professional skills training, mentoring from industry professionals and guided access to internship experiences. Trainees will be welcomed and supported regardless of race, gender, age, sexual orientation or disability. The program will facilitate this through peer mentoring, established events such as an Annual Picnic, and Poster Symposium. We will build a qCMB community through monthly meetings, and a summer symposium/retreat. Preceptors will be supported through required training that will hone their mentoring skills and convey strategies to ensure mentees engage in reproducible research. Existing and new collaborations between preceptors working in areas such as Biology of Single Cells, Post-Transcriptional Control, Genome Architecture & Function, and Protein Structure & Function will provide students with opportunities to engage in transdisciplinary quantitative/computational research that requires comprehension of cell and molecular biology. Preceptor mentoring plans and effectiveness will be evaluated by the qCMB leadership. Potential trainees (8-10 each year) are recruited into the existing interdisciplinary Cell & Molecular Biology PhD Program, and rotate during their first year to identify a lab/preceptor that fits their research interests and mentoring needs. At the end of the first year, three trainees entering qCMB labs will be selected for support through this award and each will be supported for two years before transitioning to the preceptor's research grant or other fellowship for stipend support. The award will support a total of 15 trainees, who will transition to positions in academia, or industry that require data analysis, collaboration and research experience.