Curriculum

Alan Carbajo

Machine Learning Researcher in Virology

I'm a researcher working at the intersection of artificial intelligence and virology. My main project so far: a curated, 58,046-genome dataset spanning 15 virus families, and neural networks trained on raw k-mer frequency that predict whether a virus can infect humans, then get interrogated to find out which sequence motifs actually drove that call. I currently work with Dr. Phil Pellett at Wayne State University School of Medicine in Detroit.

01 Research

What I'm chasing

Most of my work centers on one question: can you tell whether a virus infects humans just by looking at its genome? I built a dataset and a set of models to test that, and then spent most of my time trying to figure out why the models say what they say, which is where the interesting biology shows up.

My model kept flagging HHV-6 as non-human. It turns out it was picking up something real: HHV-6 integrates into host telomeres, a trick almost no other human herpesvirus uses.

01 Predicting human host-compatibility straight from raw genome sequence
02 Curating a 58,046-genome, 15-family dataset labeled for host range
03 Ablating k-mers to map which motifs actually drive a prediction
04 Tracking SARS-CoV-2's genomic drift toward human adaptation since 2019

02 Skills

Toolkit

Machine Learning & AI

ML Techniques

GANs Recurrent Neural Networks Random Forest Ensemble/Hybrid Models HMM Regression Models

Frameworks & Libraries

TensorFlow PyTorch Keras scikit-learn LangChain

Natural Language Processing

BERT Transformers State Space Models LLM Agent Design Sentiment Analysis

Computer Vision

CNN Image Segmentation Object Detection

Virology & Genomics

Molecular Virology

Virus Plaque Assays Viral DNA Isolation DNA detection with PCR

Immunology

Mammalian Cell Culture Fluorescence Imaging Antibody Staining

Bioinformatics

Sequence Analysis Genomic Data Processing

Comparative Genomics

Evolutionary Analysis Genome Annotation

Engineering

Languages

Python JavaScript R SQL Bash

Web

HTML CSS React Node.js

Cloud & HPC

Google Cloud Amazon AWS Microsoft Azure Parallel Computing GPU Acceleration

03 Publications

Published work

Sequence based virus host prediction: a curated dataset and generalizable framework for training artificial intelligence to identify viruses of humans

Alan L. Carbajo Jr., Taylor A. Vensko, Philip E. Pellett — Virus Evolution, Vol. 12, Issue 1, veag009 · Feb 2026

04 Talks