At BioComs Lab, we focus on addressing critical challenges in biology and medicine using computational and statistical approaches. Our research spans multi-omics integration, advanced machine learning, and bioinformatics-driven insights to decode complex biological processes. By collaborating with experimental researchers, we aim to bridge the gap between computational predictions and biological validation, enabling discoveries that contribute to precision medicine and understanding of diseases. Our core projects encompass methylation QTL analysis, survival analysis in oral cancers, functional analysis of oral microbiomes, and the development of cutting-edge tools for analyzing metatranscriptomic data. We are dedicated to leveraging computational systems biology to untangle the intricate interplay of genes, epigenetics, and microbial communities in health and disease.
Research Areas
Methylation QTL Analysis
Methylation quantitative trait loci (mQTL) analysis is crucial for understanding how genetic variations influence epigenetic modifications, such as DNA methylation, which play critical roles in gene regulation. DNA methylation has been linked to various diseases, including cancer, diabetes, and neurological disorders, yet only a small fraction of the genetic variants identified in genome-wide association studies (GWAS) are functionally annotated. Current data suggests that fewer than 10% of GWAS loci have definitive links to regulatory or phenotypic outcomes, underscoring a significant gap in our understanding of genetic mechanisms.
Our lab tackles this issue by integrating GWAS and methylome datasets using computational frameworks that leverage statistical and deep learning models. By identifying significant mQTLs, we aim to bridge the gap between genetic predisposition and functional outcomes, providing insights into how genetic variations contribute to complex diseases. These analyses are further validated through collaborations with experimental researchers. By elucidating these genetic-epigenetic relationships, we expect to uncover biomarkers and potential therapeutic targets, advancing the field of personalized medicine.
Survival Analysis in Oral Cancers
Oral cancer remains a significant global health issue, with approximately 377,700 new cases diagnosed annually worldwide. Despite advances in cancer research, the five-year survival rate for oral cancer patients remains below 50% in many populations, primarily due to late diagnosis and insufficient prognostic tools. While molecular profiling technologies have generated large datasets on gene expression and epigenetics, integrating these data into clinically actionable insights remains a challenge.
At BioComs Lab, we address this challenge by employing computational models to analyze multi-omics data, such as gene expression and DNA methylation profiles. Using advanced survival analysis techniques and machine learning algorithms, we identify molecular biomarkers and pathways that correlate with survival outcomes. These predictions are further validated through collaborations with experimental partners, ensuring translational relevance. Our goal is to develop predictive models that can be used by clinicians to identify high-risk patients and personalize treatment approaches, ultimately improving survival rates and patient quality of life.
Functional Oral Microbiome in Oral Diseases
The human oral microbiome is a dynamic and complex ecosystem composed of microbial species that exist in roughly a 1:1 ratio with human cells. These microbes, including bacteria, fungi, and viruses, play essential roles in maintaining oral and systemic health. Dysbiosis within the oral microbiome has been linked to diseases such as dental caries, gingivitis, and periodontitis, with significant consequences for global health. For example, periodontitis has been implicated in systemic diseases, including cardiovascular conditions and diabetes, making the understanding of its microbial drivers even more critical.
Our research focuses on understanding the functional roles of oral microbiomes in health and disease. Using metatranscriptomics, we study the active metabolic and regulatory pathways within microbial communities, uncovering their contributions to oral diseases. Our lab develops computational tools specifically designed to analyze multi-kingdom interactions, allowing for the integration of functional data from bacteria, fungi, and viruses. By identifying microbial biomarkers and pathways linked to disease states, our work paves the way for novel therapeutic strategies, such as targeted microbial interventions or personalized oral healthcare solutions.
Machine Learning and Deep Learning for Biological Data Analysis
The sheer scale and complexity of modern biological datasets pose significant challenges for traditional analytical methods. High-throughput technologies now generate massive multi-omics datasets, including gene expression, methylation, and microbiome profiles, with each dataset often encompassing millions of features and thousands of samples. Despite this, only a small proportion of the data is effectively analyzed, limiting our understanding of underlying biological mechanisms.
BioComs Lab develops and applies advanced machine learning (ML) and deep learning (DL) algorithms to tackle these challenges. For instance, our attribution frameworks for gene expression data enable us to identify key drivers of biological outcomes, providing mechanistic insights into disease processes. These methods are designed to handle high-dimensional data, uncovering patterns that would otherwise remain hidden. By applying these computational tools, we aim to enhance data interpretation, facilitate biomarker discovery, and accelerate the development of precision medicine approaches.
Tools for Metatranscriptomics Data Analysis
Metatranscriptomics provides an unprecedented view of microbial community function, offering insights into the active metabolic and regulatory processes of microbiomes. However, the complexity of metatranscriptomic data, which includes multi-kingdom interactions and vast functional diversity, presents significant analytical challenges. Existing tools are often limited in their ability to capture these interactions comprehensively, hindering progress in understanding microbial contributions to health and disease.
Our lab addresses these challenges by developing scalable, user-friendly computational tools that analyze metatranscriptomic data with a focus on multi-kingdom interactions. These tools allow researchers to uncover functional activity within microbial communities and link this activity to disease phenotypes or environmental conditions. Through collaborations with experimental researchers, we apply these tools to datasets from oral microbiomes and other ecosystems, revealing functional pathways that can inform targeted interventions. The insights gained from these analyses hold the potential to transform our understanding of microbiome dynamics and their impact on health and disease.
