Luiz Piochi

About Me

I'm a Computer Science PhD candidate at the National Institute for Research in Computer Science and Automation (INRIA). I really like using machine learning and deep learning to solve complex biological problems. Currently, I work mainly with the characterization of protein interactions and therapeutic protein design (from peptide binders to vaccines).

I believe my background in both life sciences and computer science allows me to approach research questions from a unique perspective, even in problems that do not directly involve biological data. After all, understanding biochemistry is to understand complex systems, which mirrors the intricacies of computational models.

Experience

Education

PhD in Computer Science @ INRIA & l'Université de Lorraine

September 2024 - present

Thesis: "Learning host-pathogen surface interactome to design novel therapeutics."

Summer School @ the University of Oxford

August 2025

Intensive program organized by AI for Global Goals on AI for Health & Biotech applications.

Topics: Representation Learning, Geometric Deep Learning, LLMs, Causal ML

MSc in Cell and Molecular Biology @ University of Coimbra

September 2020 - February 2023

Final grade: 19/20 (GPA: 3.9)

Thesis: "Multi-omics-based Drug Response Prediction of Cancer Cell Lines with DELFOS"

BSc in Biochemistry @ University of Coimbra

September 2017 - June 2020

Final grade: 17/20 (GPA: 3.6)

Thesis: "Sestrin2 and Mitochondrial Quality Control in Myogenic Differentiation"

Tecnólogo* in Business Administration @ ETEC

February 2014 - December 2016

* In Brazil, this type of degree exists alongside the "classical" undergraduate degree types. In Europe, it is equivalent to EQF4 degrees

Thesis: "The Importance of HR within Companies: Teamwork and Divergences"

Skills

Programming

Basic user: JavaScript

Experienced user: Python, R, SQL

Frameworks & Tools

Numeric: NumPy, pandas

ML/DL: PyTorch, Keras, scikit-learn, Hugging Face

Workflow & MLOps: Snakemake, MLflow, Weights & Biases, CI/CD

Cloud & DevOps: AWS, Docker, Git

Bioinformatics: Protein design, scRNA-seq, Metagenomics, Genomics

Languages

Native: Portuguese

High proficiency: English (C2), Spanish (C1)

Basic proficiency: French (A2), German (A1)

Publications

Portfolio

ppIRIS Image

ppIRIS

LLM-based deep learning model for predicting protein-protein interactions.

View Details
GORGON Image

GORGON

Developed a scalable and modular Snakemake pipeline for comprehensive microbial analysis, integrating tools like Kraken2 and Phyloseq for efficient processing and reporting.

View Details
DELFOS Image

DELFOS

Developed a predictive algorithm leveraging multi-omics data to identify effective cancer drug responses, validated against unseen cell lines and drugs. Published on OUP Bioinformatics.

View Details

Certifications

Awards