a Bioinformatics student at Paris-Saclay
Computational biology student interested in integrative, systems-level representation of biological data.
Interested in building new computational methods and tools merging multi-omics data from different modalities. I like understanding integrative systems behavior, particularly in human diseases and traits. I'm mostly into deep learning and network science approaches, extending my interests into single-cell and spatial omics in oncology and cognition.
Bioinformatics and Biostatistics Hub, Institut Pasteur
M1 Internship
Performed large scale bulk omics analysis on CCLE Breast Cancer data
Lebanese American University, Byblos
Undergraduate Research
Exploration of network inference approaches to model gene interactions leading to antimicrobial resistance in bacterial species. The project was later focused on E. coli with phenotypic screenings narrowed down to 11 drugs, public data retrieved from BV-BRC PATRIC db.
Microbial Genomics Lab, Lebanese American University
Undergraduate Research
Bioinformatics contributor to multiple microbial and viral genomics projects, handling WGS data from Illumina and Oxford Nanopore sequencing.
M1 grade: 16.63/20
Track: Genomics, Informatics, Mathematics for Health and Environment
Main coursework: functional and Comparative genomics, NGS analysis and statistical methods, Machine Learning and Deep Learning for precision medecine, Big Data and Databases, Advanced programming (OOP, parallel), Algorithms, Metaheuristics, Biological networks and Modeling
CGPA: 3.8/4.0
Graduted with High Distinction, Honors Program scholar, merit-based Scholarship Holder
Main Coursework: Network Science, Data mining, Genetics, Molecular Biology, Adv. Human Genetics, Biochemsitry, Algorithms and Data Structures, Linear Algebra, Honors Capstone project course
Python library for extraction, processing, visualization, and analysis of RNA structures, developed as a collaborative course project.
Genetic algorithm–based genome assembly formulated as a Hamiltonian path optimization problem, combining evolutionary computation and graph-based assembly.
Object-oriented Python implementation of a neural network library from scratch, covering core components like layers, activations, loss functions, optimizers, and training loops.
Gene–gene interaction network constructed from pangenome presence–absence data.
Big data analysis of COVID-19 epidemiological data, focusing on preprocessing, statistical exploration, and temporal dynamics.
Agent-based simulation of tumor, immune, and normal cell interactions in a 2D grid environment, replicating a published modeling study.
Julia implementation inspired by the DESeq2 package, modeling differential gene expression using negative binomial distributions and dispersion estimation.
Python library implementing generalized linear models from scratch, mimicking the R glm interface using object-oriented design principles.
Exploration of convolutional neural networks applied to brain imaging data, focusing on representation learning and model behavior.
Lightweight Python tool for animating dynamic networks through a graph class that captures node and edge changes over time.
Scripts to build a protein structure proximity network of residue nodes and distance-based edges with their visualizations and contact maps - practicing data processing into networks to further train models on these graphs
Naive motif matcher implemented in C, exploring parallel programming techniques for DNA sequence alignment with maximum Hamming distance constraints.
Genome assembly using De Bruijn and k-mer graph approaches, exploring alternative graph-based strategies for sequence reconstruction.
Association and linkage analysis of multifactorial and Mendelian diseases using multiple NGS-based analytical approaches.
Implementation of core machine learning algorithms from scratch in Python, emphasizing visualization and algorithmic understanding.
Exploration of recurrent neural networks applied to protein sequences to model aggregation-related patterns.
Hidden Markov Model implemented from scratch for splice site prediction using the Viterbi algorithm.
Optimized genetic algorithms for designing protein sequences that fold into target structures, inspired by classical GA formulations.
Exploration of graph neural networks applied to weighted gene coexpression networks derived from transcriptomic data.
Comparative study of targeted node and edge attacks based on topological properties and their effects on different network types.
Database management system for storing and querying drug discovery–related data, developed as a first project in database systems.
Collection of solutions to classic bioinformatics algorithmic problems from the ROSALIND platform.