The Mooney Laboratory uses computers and computer science to both generate new hypotheses and make biomedical research easier. To that end we are currently focusing on several areas.
1. Understanding and characterizing the molecular effects of genome variation
Over the past several years, there has been a significant amount of research activity to understand how genetic mutations alter phenotypes and give rise to disease (See Mooney 2005). Many methods now exist to understand how amino acid substitutions in proteins are disrupting protein function, and thereby giving rise to clinical phenotype (we call these ‘functional mutations’). Improvement of these methods is still an important area for research (their accuracies are usually around 65-80%). We have been working to develop new methods that can quantify the disrupted molecular function and the systems their products act upon.
To that end we have developed several resources:
- MutDB - Annotation of amino acid substitutions and single nucleotide polymorphisms with features we believe likely to affect phenotype
- MutPred - In collaboration with Predrag Radivojac at Indiana University and the Human Gene Mutation Database, we developed a new supervised machine learning method for classification of disease-causing amino acid substitutions and prediction of the underlying molecular effects of those mutations.
- In silico Functional Profiling - In collaboration with Predrag Radivojac at Indiana University, the Human Gene Mutation Database, and the National Center for Biomedical Ontology at Stanford we have developed a method for annotating the molecular effects of specific amino acid substitutions.
- Phenopred - In collaboration with Predrag Radivojac, we helped his group develop a supervised machine learning tool for predicting disease causing genes using a training set of known disease causing genes and attributes based on protein interaction, sequence and structure.
2. Characterizing the functional amino acid environments of proteins using bioinformatics
Much effort has been spent to predict the function and regulation of gene products discovered in the context of genome sequencing efforts. On aspect of this that we believe is important is prediction of functions that amino acids might participate in when a protein is expressed and operating normally. We believe understanding the scope of these functions will enable insights into human genetic diseases.
To this end we have developed several tools:
- The Structure Based Local Environment Search Tool (S-BLEST) - We developed an unsupervised approach for discovering structural relationships based on the Russ Altman's FEATURE algorithm. S-BLEST can automatically annotate environments that are statistically associated with particular functions including Gene Ontology, EC numbers, etc.
- The Catalytic Residue Predictor (CRP) - We developed a structure based predictor of catalytic amino acids based on structure.
3. Modeling the systems biology of data derived from genome wide experimentation in humans and model organisms
Currently many researchers apply a variety of experimental high throughput techniques to elucidate the genes involved in certain processes and phenotypes. These methods include proteomics approaches, ChIP-Seq, RNA-Seq, RNAi, etc. We are both collaborating with experimental researchers and developing new methods to build hypotheses on the underlying molecular explanation for observed lists of genes and/or proteins.
4. Development and support of biomedical research cyberinfrastructure
We are highly interested in building the computer networks and software the enable 21st century biomedical research. To that end, we have developed many web-based applications both to support research, enable communities of scientific researchers and to manage collaboration.
- Laboratree - Laboratree is an web based collaboration platform based on social networking (Google's OpenSocial project). Laboratree enables messaging of groups and projects, document and dataset version control and application development.
- The IndianaCTSI - We helped build and advise the Indiana Clinical and Translational Sciences Institute HUB.
- SeqMap - Working with the National Gene Vector Biorepository we have developed SeqMap, a web based tool for managing and analyzing vector integration sites within the context of gene therapy basic research.
- The National Cell Repository for Alzheimer's Disease (NCRAD) - We developed a web-based inventory management system to manage the cell, DNA and tissue samples managed by this project.