Computational virology · Structural biology · Phylodynamics

M H M Mubassir Evolution leaves a
structural signature.

I study how viral evolution changes molecular structure, receptor biology, and host range, linking global phylogenetics with structural biology.

About my work Explore research ecosystem

Ph.D. Candidate, Bioinformatics University of Georgia

37K+global H5 HA sequences

4K+unique H5 variants

1996 to 2026global evolutionary window

Scroll

00 / About

MHM

Athens, GeorgiaStructure · evolution · phylodynamics

Scientist, educator,
framework builder.

I am a Ph.D. candidate in Bioinformatics at the University of Georgia, advised by Dr. Justin Bahl, and a Lecturer at BRAC University.

My training spans agriculture, biotechnology, bioscience, pharmacy, protein design, and bioinformatics. I have worked across influenza A and B, RSV, hMPV, SARS-CoV-2, mpox, and plant immune receptors. This experience shaped how I approach viral emergence as a connected problem across evolution, structure, molecular mechanism, ecology, and computation.

CurrentBahl Lab · UGA

Ph.D. expected2027

TeachingTop 1% evaluation

TrainingBioinformatics · Pharmacy · Bioscience

Outside science I enjoy filmmaking and visual storytelling.

View publications on Google Scholar

01 / Scientific vision

From genetic diversity
to functional risk.

Rather than studying mutations in isolation, I investigate how standing phenotypic diversity emerges across whole viral populations, and how that diversity shapes future host transitions.

My long-term goal is to identify functional evolutionary change before it becomes visible through outbreaks, enabling earlier surveillance, improved risk assessment, and more proactive pandemic preparedness.

01Genetic diversityPopulation variation

02Structural consequenceReceptor biology

03Functional breadthStanding phenotype

04Cross-host riskEarlier signal

02 / Research ecosystem

Integrating evolution,
structure, and ecology.

Drag any bubble · tap a domain to focus · select a tool to inspect

Research map

How I study viral evolution

Explore the connected biological questions, methods, and software that link viral evolution with structure, ecology, and preparedness.

Research ecosystem

03 / Structural biology

Structure,
in real space.

An experimental H5 hemagglutinin structure anchors the structural side of my research, connecting receptor recognition, molecular dynamics, energetic interpretation, and evolutionary change.

ProteinH5 HA trimer

VirusA/Texas/37/2024

LigandAvian analog LSTa

Resolution2.32 Å

Drag to rotate · scroll to zoom · open the PDB record for full experimental details.

Structure source PDB 9DIP, reported by Lin et al. in Science 386, 1128 to 1134 (2024). Paper PDB record

Experimental structure 9DIP

Experimental structure

H5 hemagglutinin trimer

A/Texas/37/2024 (H5N1) · bound to the avian receptor analog LSTa

X-ray diffraction2.32 Å PDB 9DIP

04 / Research visualization · Phylodynamics

Evolution,
seen in motion.

This animated phylogeny introduces the evolutionary history of U.S. H5N1 clade 2.3.4.4b. Play the timeline here, then open the full Nextstrain build to explore lineage structure, host transitions, geography, and phenotype-aware annotations.

Open full Nextstrain Animated overview · live tree opens separately

PhylogenyHostStructurePhenotype

Animated phylogeny of U.S. H5N1 clade 2.3.4.4b

Evolutionary timeJan 2021

Explore live tree

13K+U.S. HA sequences

2,195unique HA variants

Host-awareavian · cattle

Phenotype-awaresequence · structure · function

05 / Current research program

Two current projects,
one future direction.

A connected program that moves from population-scale H5 evolution to structure-informed prediction.

Aim 1 · United States · 2021 to 2025

A population-scale phenotype atlas of H5N1 clade 2.3.4.4b

What functional capacity was already present before and during emergence in cattle?

This project integrates approximately 13,000 U.S. hemagglutinin sequences and 2,195 unique HA variants with a time-resolved phylogeny, AlphaFold 3 structures, receptor-binding-site geometry, physicochemical traits, Rosetta and FoldX energetics, molecular dynamics simulations, and glycan-binding interpretation. It tests how standing phenotypic breadth, lineage sorting, and distributed structural change shape cross-host potential.

First-author manuscript under review at Nature

13K+ HA sequences2,195 variantsAvian and cattle lineages

Sequence
Structure
Function
AlphaFold 3
Rosetta
Molecular dynamics
Nextstrain

Aim 2 · Global H5 · 1996 to 2026

A global evolutionary and phenotypic atlas of H5

How do ecological opportunity, host movement, and phenotypic breadth combine to shape the global evolutionary history of H5?

Using more than 37,000 H5 hemagglutinin sequences collected from 1996 through 2026, this aim reconstructs how genetic and phenotypic diversity is generated, transmitted, and retained across hosts, regions, clades, and epidemic waves. Time-scaled phylogenies and phylogeographic models quantify lineage movement, host transitions, persistence, and reassortment. Markov jumps and rewards estimate cross-host flow and residence, while structure-derived receptor-binding, stability, and physicochemical traits map functional change onto the tree. The objective is to identify repeatable evolutionary phases and lineages where ecological opportunity converges with phenotypic breadth, creating priority targets for surveillance.

Global dataset curated through 2026

37K+ HA sequences4K+ variants1996 to 2026

Phylogeography
Host transitions
Markov jumps and rewards
Lineage persistence
Reassortment
Structural phenotype

Future direction · Predictive evolutionary biology

A structure-informed predictive framework for phenotype and lineage fate

Can evolutionary history, protein structure, host ecology, and measured phenotype be learned together rather than analyzed as separate data streams?

This future program will develop statistical and machine-learning frameworks that jointly represent sequence variation, time-scaled phylogenies, protein structure, receptor-site geometry, phenotype, host transitions, and ecological context. The conceptual architecture combines protein language models with structure-aware encoders, multimodal fusion, and flow matching to trace phenotype evolution, rank lineage fate, and identify surveillance priorities. RFdiffusion provides a reference for generative structural modeling, while the central methodological goal remains structure-informed phylodynamics. Foundational training and pilot design are in progress.

Foundational training and pilot design in progress

Future directionJoint evolutionary modelPilot studies planned

Protein language models
Flow matching
Multimodal learning
Structure-informed phylodynamics
RFdiffusion

06 / Earlier platforms

Built across
multiple viruses.

My structural-biology foundation began in computational vaccine and antigen design, before expanding into population-scale viral evolution.

AM.S. thesis · Influenza H3N2

Consensus-based influenza vaccine design

Designed and evaluated consensus HA immunogens to capture sequence diversity while preserving structurally and antigenically important regions, using comparative modeling and integrated sequence and structure analysis.

Consensus designHAAntigen engineering

BStrauch Lab · Viral fusion proteins

A Rosetta-guided proline scanner

Developed a computational design workflow that introduced candidate proline substitutions, evaluated structural consequences, compared energetic scores, and prioritized mutations for prefusion stabilization across influenza, RSV, hMPV, and SARS-CoV-2 systems.

RosettaPrefusion stabilizationProline scanning

CRespiratory-virus structural biology

Multi-virus protein and antigen modeling

Applied structure prediction, energetic analysis, docking, and molecular dynamics simulation to questions in receptor specificity, glycosylation, antigen conservation, and vaccine design across respiratory viruses.

RSVhMPVSARS-CoV-2Influenza A/B

M H M Mubassir pipetting samples at the bench

M H M Mubassir operating laboratory instrumentation

Experimental foundations

Computational questions grounded in basic wet-lab experience.

I have foundational hands-on experience with DNA extraction and purification, PCR, agarose gel electrophoresis, marker-assisted selection, basic cell-culture training, and protein purification. This background supports how I interpret molecular, structural, and experimental data without overstating my role as a primarily computational scientist.

DNA extractionDNA purificationPCRGel electrophoresisMarker-assisted selectionCell-culture basicsProtein purification

Influenza AInfluenza BRSVhMPVSARS-CoV-2Mpox

Mubassir, M. H. M.; Nasiruddin, K. M.; Shahin, N. H.; Begum, S. N.; Saha, M. K.; Rashid, A. Q. M. B.

Journal of Pharmaceutical, Chemical and Biological Sciences 4(4), 475–486

2016

Published

Morpho-molecular screening for bacterial leaf blight resistance in some rice lines and varieties

Mubassir, M. H. M.; Nasiruddin, K. M.; Shahin, N. H.; Begum, S. N.; Saha, M. K.; Rashid, A. Q. M. B.

Journal of Plant Sciences 4(6), 146–152

08 / Open science links

Tools to use.
Ideas to learn.

A small, curated set of browser tools and learning resources that complement the research program.

Web toolkit

Open scientific platforms

01Structure predictionAlphaFold ServerGoogle DeepMind 02Structure predictionRobetta / RoseTTAFoldBaker Lab 03Biomolecular AIBoltzMIT · Recursion 04Accessible foldingColabFoldColabFold 05Experimental structuresRCSB Protein Data BankRCSB PDB 06PhylogenomicsNextstrain / AuspiceNextstrain

Learning shelf

Talks, courses, and communities

01Protein design communityBoston Protein Design and Modeling ClubBPDMC 02AI protein designMachine Learning for Protein Design BootcampRosetta Commons 03Structural biologyProtein folding explainedGoogle DeepMind 04PhylodynamicsPhylodynamics: concepts and applicationsFogarty International Center 05Genomic surveillanceNextstrain narratives tutorialNextstrain 06Machine learningCS229: Machine LearningStanford Online

Bayesian phylogenetics

Hands-on tutorials: Taming the BEAST

09 / Conferences

Ideas in public.

Selected posters and presentations on influenza evolution, receptor binding, molecular simulation, and protein structure.

M H M Mubassir presenting research at a scientific meeting — Presenting research on viral evolution and structural biology.

Conference presentation

An evolutionary trajectory of H5 hemagglutinin phenotypes across U.S. clade 2.3.4.4b.

Ecology and Evolution of Infectious Diseases · Virginia Tech · Blacksburg, Virginia

2026Presentation

An evolutionary trajectory of H5 hemagglutinin phenotypes across U.S. clade 2.3.4.4b

Ecology and Evolution of Infectious Diseases · Virginia Tech · Blacksburg, VA · June 1 to 4

2026Presentation

Phenotypic flexibility, rather than optimization, governs pandemic potential

CEIRR 5th Annual Network Meeting · Emory University · Atlanta, GA · May 26 to 29

2025Poster

An integrative atlas of receptor-binding evolution in U.S. H5N1 2.3.4.4b viruses

CEIRR 4th Annual Network Meeting · University of Georgia · Athens, GA · May 27 to 30

2024Poster

Integrating molecular dynamics and phylodynamics to unravel sialic-acid and HA interaction patterns in avian H5 2.3.4.4b influenza

CEIRR Conference · New York, NY

2022Poster

Establishing a proline scanner for stabilizing virus class I fusion proteins

RosettaCon · USA

2017Presentation

Structural dynamics of endogenous peptide AtPep1

ASIA International Multidisciplinary Conference · Malaysia

10 / Science in practice

Scientific exchange,
beyond the screen.

Research grows through shared ideas, thoughtful conversation, and engagement with the wider scientific community.

M H M Mubassir discussing protein structure with a colleague — Protein structureScientific exchange around protein modeling, design, and structural biology.

M H M Mubassir in conversation with a colleague about predictive biology — Predictive biologyCommunity conversations on emerging methods in AI and protein science.

11 / News

Recent signals.

A concise record of recent publications, public research, and milestones.

JUL 2026Manuscript

First-author H5N1 phenotype manuscript is under review at Nature.

JUN 2026Publications

Let’s connect across
evolution, structure, and phylodynamics.

For research discussion, collaboration, invited talks, or academic correspondence.

mhmmubassir@uga.edu

M H M Mubassir Evolution leaves astructural signature.

Scientist, educator,framework builder.

From genetic diversityto functional risk.

Integrating evolution,structure, and ecology.

How I study viral evolution

Structure,in real space.

H5 hemagglutinin trimer

Evolution,seen in motion.

Two current projects,one future direction.

A population-scale phenotype atlas of H5N1 clade 2.3.4.4b

A global evolutionary and phenotypic atlas of H5

A structure-informed predictive framework for phenotype and lineage fate

Built acrossmultiple viruses.

Consensus-based influenza vaccine design

A Rosetta-guided proline scanner

Multi-virus protein and antigen modeling

Computational questions grounded in basic wet-lab experience.

Selected work.

Standing HA phenotypic breadth shapes H5N1 cross-host potential.

Standing HA phenotypic breadth shapes H5N1 cross-host potential

Dynamic Risk Maps Predict Highly Pathogenic Avian Influenza Hotspots Across North America

Ecological and evolutionary drivers of 2.3.4.4b H5Nx HPAI spread in Europe

A phylogeny-informed mathematical modeling of HPAI H5N1 transmission dynamics and effectiveness of control measures

A time-delayed epidemic model for control of highly pathogenic avian influenza (HPAI H5Nx) with vaccination, compliance, and non-pharmaceutical interventions

N-glycosylation at the receptor binding site drives differences in receptor binding specificity between influenza B virus lineages

Evolving fitness and immune escape: A retrospective analysis of SARS-CoV-2 spike protein (2020–2024) using a protein language model

MHC-II-EpiPred: Sliding-window scanning for MHC class II epitope discovery via a protein language model

PhyloSpillover: Graph neural network-based prediction of pre-emergence zoonotic spillover risk from avian influenza H5N1 using multi-host phylodynamic simulations

A graph neural network framework for epidemiologically informed phylodynamics of avian influenza H5N1 reassortment in Europe

Phenotypic diversity as an early signal of zoonotic spillover risk

Host lineage shapes gut microbiome structural architecture at the poultry–wildlife interface of avian influenza

A reproducible R-based protocol for preprocessing avian influenza A (H5Nx, clade 2.3.4.4b) sequence data from GISAID for phylogenetic analysis

Standardized wild and domestic host annotations for global avian influenza A (H5Nx, clade 2.3.4.4b) sequences: A curated classification resource

Modeling the risk of HPAI H5N1 spread through cattle networks: Implications for cross-species transmission and pandemic prevention

Quantifying cross-host transmission and seasonal phase structure of U.S. H5N1 using integrated phylodynamics and incidence

A mathematical model for understanding and controlling monkeypox transmission dynamics in the United States and its implications for future epidemic management

Effect of D128N mutation in OsSERK2 in Xa21-mediated innate immunity

How Arabidopsis receptor-like kinase 7 (RLK7) manifests: Delineating its structure and function

Comprehensive in silico modeling of multi-domain plant PRR Xa21 and its interaction with PAMP RaxX21-sY

A brief overview on early events of Xa21-mediated pattern-triggered immunity

A synopsis of different plant LRR-RLK structures and functionality

In-silico structural modeling and molecular dynamics simulation of pathogen-associated molecular pattern RaxX21

Measurement of phenotypic variation for control and bacterial-leaf-blight-inoculated rice lines and varieties

SSR marker-based genetic diversity analysis of some rice lines and varieties for bacterial leaf blight resistance