Marcelo Der Torossian Torres
Peptides · Antibiotics · AI

Marcelo Der Torossian Torres

Research Assistant Professor · University of Pennsylvania
Machine Biology Group · Bioengineering & Psychiatry

I am a Research Assistant Professor at the University of Pennsylvania in the Machine Biology Group, working where peptide chemistry, microbiology, and artificial intelligence meet. My work designs and discovers new antibiotics to confront the global rise of antimicrobial resistance.

Trained as a synthetic and organic chemist, I pair solid-phase peptide synthesis and structure–activity-relationship-guided design with deep learning and generative AI to mine the proteomes of living, extinct, and microbial life for encrypted antimicrobial peptides. I also engineer peptide chemical modifications for stability and selectivity, and build low-cost biosensors for rapid infectious-disease diagnostics.

Affiliations

News

Dec 2025
Began as Research Assistant Professor at the University of Pennsylvania (Depts. of Bioengineering & Psychiatry).
2026
A generative artificial intelligence approach for antibiotic optimization published in Nature Machine Intelligence.
2026
Deep learning reveals antimicrobial peptides within prions published in Nature Microbiology.
2025
Discovery of antibiotics in the archaeome using deep learning published in Nature Microbiology.
2024
Two Cell papers mine the human gut and the global microbiome for new peptide antibiotics.
01

Antimicrobial peptides & SAR-guided design

Dissecting how physicochemical and structural descriptors — charge, hydrophobicity, amphipathicity, helicity — govern the activity, selectivity, and stability of cationic antimicrobial peptides, and using those rules to redesign natural toxins (decoralin, polybia-CP, VmCT1, mastoparans) into potent, non-toxic synthetic therapeutics.

02

AI & deep learning for antibiotic discovery

Building computational and generative-AI pipelines — pattern recognition, genetic algorithms, deep learning, and diffusion/Bayesian models — to uncover encrypted antibiotics across the tree of life: the human proteome, extinct organisms (molecular de-extinction), the archaeome, and the global microbiome.

03

Peptide chemical modification

Expanding peptide chemical space with cyclization, disulfide and lactam bridges, non-canonical residues, backbone and terminal modifications, and computationally designed pore-forming scaffolds — tuning proteolytic stability, membrane selectivity, and function.

04

Low-cost biosensors & diagnostics

Designing inexpensive electrochemical and colorimetric biosensors that detect bacterial and viral pathogens in minutes at the point of care — from pencil-graphite SARS-CoV-2 tests to paper-based devices for antibiotic-resistant bacteria.

Selected publications

Full list (98) →
★Torres, MDT; Wan, F; de la Fuente-Nunez, C. Deep learning reveals antimicrobial peptides within prions. Nature Microbiology, 2026.
DOIAI & deep learningAMP design & SAR
★Torres, MDT; Zeng, Y; Wan, F; Maus, N; Gardner, J; de la Fuente-Nunez, C. A generative artificial intelligence approach for antibiotic optimization. Nature Machine Intelligence, v. 8, p. 841-856, 2026.
DOIAI & deep learning
★Torres, MDT; Wan, F; de la Fuente-Nunez, C. Discovery of antibiotics in the archaeome using deep learning. Nature Microbiology, 10, 2153-2167, 2025. In Press.
DOIAI & deep learning
★Torres, MDT; Brooks, E; Cesaro, A; Sberro, H; Nicolaou, C; Bhatt, AS; de la Fuente-Nunez, C. Human gut metagenomic mining reveals an untapped source of peptide antibiotics. Cell, v. 187, i. 19, p. 5453-5467.E152024.
DOIAI & deep learningAMP design & SAR
★Santos-Junior, CD; Torres, MDT; Duas, Y; del Rio, AR; Schmidt, TSB; Chong, H; Fullam, A; Kuhn, M; Zhu, C; Houseman, A; Somborski, J; Vines, A; Zhao, XM; Bork, P; Huerta-Cepas, J; de la Fuente-Nunez, C, Coelho, LP. Computational exploration of the global microbiome for antibiotic discovery. Cell, v. 187, i. 14, p. 3761-3778.E16, 2024.
DOIAI & deep learning
★Wan, F; Torres, MDT; Peng, J, de la Fuente-Nunez, C. Molecular de-extinction of antibiotics enabled by deep learning. Nature Biomedical Engineering, v.8, i. 7, p. 854-871, 2024.
DOIAI & deep learning
★Torres, MDT; Melo, MCR; Flowers, L; Crescenzi, O; Notomista, E; de la Fuente-Nunez, C. Mining for encrypted peptide antibiotics in the human proteome. Nature Biomedical Engineering, v. 6, i. 1, p. 67-75, 2022.
DOIAI & deep learningAMP design & SAR
★Silva, ON; Torres, MDT; Cao, J; Alves, ESF; Rodrigues, LV; Resende, JM; Lião, LM; Porto, WF; Fensterseifer, ICM; Lu, TK; Franco, OL; de la Fuente-Nunez, C. Repurposing a peptide toxin from wasp venom into anti-infectives with dual antimicrobial and immunomodulatory properties. Proceedings of the National Academy of Sciences of the United States of America, v. 117, i. 43, p. 26936-26945, 2020.
DOIAMP design & SAR