Computational Biology

Overview

Decoding Biological Complexity Through Data Science and Algorithms

Computational biology merges computer science, quantitative mathematics, and biological principles to analyze vast biomedical datasets and construct mechanistic models of living systems. It forms the computational backbone of modern genomics, precision medicine, structural biology, and targeted drug discovery.

In this program, students transition from theoretical algorithms to real-world bioinformatics pipelines and systems modeling frameworks. By analyzing high-throughput genomic data, engineering computational workflows, and interpreting dynamic biological networks, participants cultivate strong computational thinking and data-driven intuition for academic research and biotech industry careers.

Selected Topics

Sequence Alignment & Comparative Genomics

Focuses on sequence alignment methods, including dynamic programming approaches such as Needleman–Wunsch and heuristic tools such as BLAST to align nucleic acid sequences and identify evolutionary conservation.

Protein Structure Prediction & Molecular Docking

Leverages deep learning models and docking tools like AlphaFold and AutoDock to predict protein structures and model ligand–protein interactions.

Phylogenetic Reconstruction & Evolutionary Trees

Applies maximum likelihood and Bayesian statistical methods to build evolutionary trees and trace ancestral lineages across biological species.

Transcriptomic Pipelines & RNA-Seq Analysis

Processes high-throughput sequencing reads through quality filtering, pseudoalignment, and differential expression statistical testing.

Metabolic Flux Analysis & Constraint-Based Modeling

Models genome-scale metabolic networks using linear programming and Flux Balance Analysis (FBA) to predict steady-state cellular phenotypes.

Molecular Dynamics Simulations & Biophysical Forcefields

Simulates physical atomic movements over time to study protein conformational dynamics, folding pathways, and thermodynamic stability.

Sample Project Design

Gene Regulatory Network Modeling & Dynamic Perturbation Simulation

In this project, students reverse-engineer gene regulatory networks from multi-condition and time-series gene expression datasets using statistical inference algorithms and ordinary differential equations (ODEs).

Research Question

How do transcription factor interactions dynamically govern downstream gene regulatory networks during cellular perturbation?

Computational & Project Activities

Students execute analytical workflows and modeling pipelines, including:

Preprocessing raw RNA-seq expression matrices via normalization and log-transformation.
Inferring pairwise gene co-expression and regulatory links using LASSO regression and Mutual Information.
Constructing directed network graphs and calculating node centrality, clustering coefficients, and hubs.
Formulating systems of Ordinary Differential Equations (ODEs) to capture time-series regulation kinetics.
Simulating targeted gene knock-outs and overexpressions in silico to predict network responses.
Benchmarking predicted regulatory links against curated databases and literature resources such as TRRUST, with STRING used for complementary functional association analysis.

Data Analysis & Evaluation

Transforming computational models into biological insights:

Network Topology & Graph Mining Students identify scale-free properties, modular sub-networks, and master regulatory hubs using NetworkX and Cytoscape.

Dynamic Perturbation Modeling Participants simulate steady-state shifts and feedback loops under simulated genetic knock-out conditions.

Benchmarking & Sensitivity Analysis Students evaluate model accuracy using Precision-Recall curves against gold-standard DREAM Challenge datasets.

Industry Context & Translational Research

Gene regulatory network modeling lies at the core of computational drug discovery, target validation, and synthetic biology applications in modern biotech enterprises.

Target Identification Pinpointing key driver genes and master regulators to discover high-efficacy therapeutic drug targets.

Synthetic Genetic Circuits Engineering synthetic gene circuits with predictable logic gate responses for cell therapy applications.

In Silico Drug Screening Simulating multi-target drug interventions to mitigate off-target toxicity and drug resistance pathways.

Biomarker Discovery Extracting co-regulated gene modules as diagnostic panel markers for complex disease states.

Possible Project Outputs

Depending on individual progress, deliverables may include:

Reproducible Jupyter / R Markdown analytical code repository
High-resolution network graphs and dynamic simulation plots
Formal research paper formatted for computational biology journals
Interactive dashboard (Streamlit/Shiny) displaying inferred networks
Research poster detailing key regulatory drivers and biological pathways

Experimental Translation

Computational modeling connects data science with laboratory experimentation. Students explore:

How predictive modeling narrows down candidate targets and streamlines experimental validation.
How network models guide combination therapy design in oncology and immunology.
How algorithmic workflows integrate with AI-driven biology platforms in modern R&D.