AI4Science Atlas

ML for protein surfaces; deep learning protein & vaccine design

Stated pillar "ML approaches for study of protein surfaces"; MaSIF; DL pipeline for protein binder design (2024–25)

Generative ML for metabolic kinetic modeling

RENAISSANCE AI tool; Choudhury et al., Nat. Catalysis 2024

Deep learning for de novo protein & membrane-protein design

Two bioRxiv 2025 papers on DL-based protein/membrane design

Geometric deep learning for protein structure, interfaces & cryo-EM

PeSTo (Nat. Commun. 2023); PeSTo-Carbs (JCTC 2024)

ML for regulatory genomics & chromatin accessibility; generative AI for DNA design

ChromatinHD (Nat. Commun. 2025); DNA-Diffusion

ML decoders for brain–spine interfaces

Lorach et al., Nature 2023; Tabulae Paralytica, Nature 2024

Cryo-EM + ML for antibody/epitope prediction

"Decoding epitope immunodominance in HIV Env using cryoEM and ML" (2026)

Neuromorphic/ML-hardware via bio-nanopores; deep learning for real-time microscopy

ERC "NEON" grant (2025); DL-assisted SICM imaging (Small Methods 2025)

Circumstantial ML: teaches "ML for Physicists", presents at ML-in-comp-bio venues

No specific ML paper pinned down — core work is stochastic/dynamical modeling

PI’s prior work used ML for protein–protein interaction engineering; lab brand-new at EPFL

"Deploying synthetic coevolution and ML to engineer PPIs", Science 2023 (pre-EPFL)

ML-accelerated catalyst discovery; GA+ML inverse design

"ML meets volcano plots: computational discovery of cross-coupling catalysts", Chem. Sci. 2018

Transformers for reaction prediction & retrosynthesis; LLMs for chemistry

Molecular Transformer; DRFP reaction fingerprinting

ML-predicted NMR chemical shifts for de novo crystal structure determination

ShiftML / ShiftML4; JACS 2022

Explainable AI (SHAP) applied to lignin-extraction process chemistry

SHAP-based explainable ML models for lignocellulosic biomass extraction

Data-driven / ML prediction of colloidal nanocrystal synthesis outcomes

"Holistic Data-Driven Approach to Synthesis Predictions of Nanocrystal Shapes", JACS 2024–25

ML (gradient-boosted trees) for topological-materials discovery & classification

"Detection of Topological Materials with Machine Learning" (~90% accuracy)

Deep learning & statistical mechanics for protein folding and surface prediction

Lab page: "Machine learning approaches for the study of protein surfaces"

Machine-learning interatomic potentials (MLIPs) for atomistic simulation

PET-MAD, #1 on Matbench Discovery (2025)

AI-driven discovery of nanoporous materials (MOFs, zeolites)

MOFTransformer; Digital Discovery 2024

ML-augmented cluster expansions for multicomponent-alloy energetics

npj Computational Materials 2025

AI/data-driven high-throughput materials discovery; Materials Cloud/AiiDA

IBM Faculty Award for AI-curated materials datasets (2024–25)

ML for error/uncertainty propagation in first-principles materials simulation

Algorithmic differentiation for plane-wave DFT (arXiv 2025); DFTK.jl

Data-driven ML mechanics for tribology and materials damage

"ML for computational fracture and damage mechanics", Eng. Fracture Mechanics 2026

Physics-informed / scientific ML for mechanics of material failure

Lab explicitly named "Data-Driven Mechanics"; "Scientific ML" research tag

Deep learning for electron-microscopy image analysis of nanomaterials

Fritz Haber Institute talk: "Discovering Materials Transformations...Using Deep Learning"

Neural-network quantum states for many-body quantum physics

"Solving the Quantum Many-Body Problem with ANNs", Science 2017; NetKet

Quantum ML theory — trainability & barren plateaus in variational quantum algorithms

"Barren Plateaus Preclude Learning Scramblers" (2020)

Neural-network (RBM) representations of quantum states for dissipative dynamics

"Neural-Network Approach to Dissipative Quantum Many-Body Dynamics", PRL 2019

ML / protein language models for coevolution-based inference in biophysics

"Generative power of a protein language model trained on multiple sequence alignments"

Deep learning for astrophysical data — gravitational lensing, cosmology

"Deep Learning in Wide-field Surveys: Fast Analysis of Strong Lenses" (2019); ODISSEE project

Reinforcement learning for HVAC / building climate control

EPFL news: RL-trained smart controllers for indoor climate

ML (genetic algorithm) for flow control & wind-turbine blade-pitch optimization

Le Fouest & Mulleners, Nature Communications 2024

Deep learning for river-morphodynamics prediction

Lab "AI for River Morphology" project, satellite-imagery DL models

ML for industrial process & energy-system optimization

"Improved Waste Heat Management...Using a Machine Learning Approach"

Deep learning / neural nets for solid oxide cell (SOFC/SOEC) characterization

Lab site project (moderate, project-level evidence)

Neural-network-inspired data-driven models for power-converter design

Profile reference to NN-inspired models for transformer design (moderate)

ML-enhanced optimization of power distribution & grid storage

Lab description: "machine-learning-enhanced optimization of power distribution systems"

Deep learning for solar-irradiance forecasting

CNN-with-memory hybrid DL model on EPFL campus webcam imagery

Hybrid NWP+ML wind/solar forecasting; deep learning for wind-farm wake modeling

Stated lab pillar; "Offshore wind farm wake modelling using deep feed-forward NNs"

ML for hydropower turbine start-up trajectory optimization

Muser et al., "Fatigue damage reduction in hydropower startups with ML", Nat. Commun. 2025

ML / reduced-order surrogate modeling for reactor UQ & data assimilation

Lab-wide stated pillar (shared across 3 members, not individual papers)

ANN regression for gas/fluid-film bearing design; turbomachinery optimization

Iseli & Schiffmann ANN bearing-force paper; DARTS-NETGAB

ML / data-driven modeling of human–building interaction & occupant behavior

"OccuVAE: unsupervised occupancy inference in data-driven energy modeling" (2023)

Deep learning & physics-enhanced GNNs for fault diagnostics/prognostics

PHM Society Fellow (2023); "From physics to machine learning and back"

CNN-based crack segmentation; AI for seismic damage assessment

"Uncertainty quantification for a DL model for image-based crack segmentation" (2024)

ML/RL for autonomous vehicles & mobility-service optimization

Stated research: "game theory and machine learning in transportation"

Deep learning on 3D point clouds for structural damage/capacity prediction

"STEEL-3dPointClouds" dataset paper, Scientific Data (Nature, 2025)

Data-driven ML for geomechanics & subsurface-engineering upscaling

Ongoing ML project for geological carbon-storage upscaling

Deep learning / CV for pedestrian trajectory forecasting & autonomous driving

"Social LSTM"; "Human Trajectory Forecasting in Crowds", IEEE T-ITS 2021

ML for adaptive & autonomous robotic construction

Keynote title only — no paper/pillar text found

Deep learning / CV for remote sensing, ecology & biodiversity monitoring

"SSL4Eco" (2025); "ML for wildlife conservation", Nat. Commun.

Photogrammetry fused with ML/deep learning for 3D mapping

AI-based multi-scan correspondence detection, ISPRS J. Photogrammetry & RS

Statistical / ML methods for FTIR aerosol spectral interpretation

Combined statistical + ML estimation of organic functional groups from FTIR spectra

ML / deep learning for species-distribution modeling & landscape genomics

R.SamBada pipeline; "Scaling species distribution models with deep learning"

Neural-network classification of hydrometeors/snowflakes from radar & camera data

"Vertical structure of radar reflectivity...neural networks in the Swiss Alps"; MASC classifier

Graph machine learning for urban climate / microclimate forecasting

"Hourly Urban Air Temperature Forecasting with Graph Machine Learning" (2026)

Deep learning / CNNs for computational imaging through multimode optical fibers

CNN-based image reconstruction through scrambled multimode-fiber output

Deep learning / generative AI for molecular & cancer biology

Lab pillar: "AI in Molecular Medicine" (housed at ISREC)

Deep learning for computational biology (cell-type annotation, single-cell genomics)

Stated pillar: "AI applications in computational biology"

Network analysis / ML for epidemiological modeling; DL for plant disease

PlantVillage deep-learning disease detection

Biologically grounded neural-network models of learning in the brain

Long-standing pillar; standard computational-neuroscience textbooks

Deep learning for neural dynamics, proprioception & motor control

DeepLabCut (Nat. Neurosci. 2018); "Task-driven NN models predict proprioception", Cell 2024

Deep learning for animal-behavior tracking & joint neural-behavioral analysis

DeepLabCut; "Learnable latent embeddings for joint behavioural and neural analysis", Nature 2023

Computer vision / deep learning applied to biomedical imaging

Stated pillar, weaker/less specific evidence

Deep learning + biomechanical simulation to reverse-engineer insect neuroscience

DeepFly3D (2019); NeuroMechFly v2 digital-twin fly model (2024)

Computational neuroscience / ML models of animal locomotion

Lab lists "computational neuroscience...and machine learning" as core disciplines

Scientific ML & data-driven model reduction for physical simulations

"Breaking the Kolmogorov Barrier with Nonlinear Model Reduction"

ML-based reduced-order modeling for plasma-physics simulation

"Data-driven model order reduction for accelerating boundary plasma turbulence simulations"

Bio-inspired AI for robotics & computational biology

Stated pillars: "bio-inspired artificial intelligence, computational biology"

Neural / deep-network models of human visual perception

"Capsule networks as recurrent models of grouping and segmentation", PLOS Comp. Biol.

Neuromorphic / ML hardware for neuroengineering

Stated pillars: ML hardware, neural interfaces, neuroengineering

Computer vision / ML for robot perception

Stated pillars

ML for adaptive control & human–robot interaction

Stated pillars

ML / CV for robot perception & audio processing

Stated pillars

Deep learning for inverse problems in medical image reconstruction (X-ray/CT/MRI)

IEEE Trans. Image Processing 2017 (2019 IEEE SPS Best Paper Award)

Deep learning for accelerated / robust MRI reconstruction

DL-based coil-sensitivity / k-space reconstruction projects

ML generative frameworks for antenna geometry design

Conditional VAE for antenna design (Springer, 2025)

Neuromorphic hardware for biomedical / implantable neural signal processing

Invited talk: "Neuromorphic and Unconventional Computing...", 2023

Reinforcement-learning theory for renewable energy, transportation & robotics

Stated pillars

Embedded / edge ML for biomedical signal processing & wearable health monitoring

COUGHVID crowdsourced cough-ML dataset; federated learning for seizure detection

ML / deep learning for EDA — logic-synthesis automation

"Deep Learning for Logic Optimization Algorithms" (logic opt. as MDP via deep RL)

2D-material neuromorphic / logic-in-memory hardware for AI

MoS₂-based artificial-neuron circuits; "Logic-in-memory based on an atomically thin semiconductor"

Deep learning for super-resolution / single-molecule microscopy

Unsupervised DL framework decoding bacterial cell-cycle from super-resolved images

Photonic / optical neural-network hardware (neuromorphic computing)

"Photonics and AI" (Advanced Photonics interview, 2024)

Deep learning for neural / motor decoding

"Robust and accurate decoding of hand kinematics...using deep learning"

Data-driven / generative ML for turbulence & PDE solution discovery

Generative diffusion model discovering unknown Navier–Stokes periodic-orbit solutions

Machine-learned interatomic potentials for materials & mechanics

"Machine learning for metallurgy" series — NN potentials for Mg, Zr

Physics-informed ML for reduced-order modeling of PDEs

"Physics-informed machine learning for reduced-order modeling of nonlinear problems"; DeepBND

Data-driven / ML models for nonlinear mechanics of metamaterials

Public dataset released for data-driven/ML design of bistable metamaterials

AI-guided computational design for soft / agricultural robots

RoboGrammar; LLM-guided design pipeline for produce-picking soft robots

Deep learning for microscopy image analysis & centriole detection

CenFind, BMC Bioinformatics 2023

Computational / generative single-cell genomics; dynamical modeling of cell fate

RNA velocity of single cells, Nature 2018

Graph neural networks applied to genomics, transcriptomics & structural biology

REGEN (gene-network GNN); RiboGL; LD-FPG protein-structure GNN

Algebraic topology as ML-classifier input for neuron morphology (with Blue Brain)

Topological classification of pyramidal-neuron shapes feeding ML algorithms

Computational / ML modeling of single-cell & spatial epigenomic trajectories

Spatially resolved epigenomic profiling of brain organoids (bioRxiv 2022)

ML at the intersection of host–pathogen genomics and precision medicine

Genome-to-genome framework; ML genomic susceptibility markers

ML pipeline for quantifying motor behavior in disease models

"High throughput ML pipeline...zebrafish motor behavior", bioRxiv 2025

Deep-learning-enabled super-resolution imaging

"Fast DNA-PAINT imaging using a deep neural network", Nat. Commun. 2022

Machine learning for neural-prosthetic sensory encoding

"A machine learning framework that encodes images like a retina" (2024)

ML on brain graphs / fMRI for decoding brain states

"Machine Learning with Brain Graphs"; "Decoding Brain States from fMRI Connectivity Graphs"

Deep learning for medical image analysis

DL diffusion-MRI lymph-node segmentation; DL for ultrasound imaging

Deep learning for cryo-EM structure determination

Siamese neural network for cryo-EM projection-angle recovery

Deep-learning-based large-scale EM segmentation for connectomics

Algorithmic / DL approaches to 3D EM neuron segmentation

Blue Brain Project — topology + ML-based neuron classification

Objective algebraic-topology/ML classification of cortical neuron morphologies

Computational modeling, ML & data science for brain disorders

Krembil Centre for Neuroinformatics

Deep learning for aligning / registering large-scale brain datasets

Blue Brain deep-learning gene-expression dataset registration tool

ML + connectomics for clinical outcome prediction

ML system combining connectome data to predict stroke recovery

Deep learning for cell/organism image analysis & neural tracking

ML yeast-cell segmentation/tracking; NN pipeline for 4D C. elegans neuron tracking; co-author of CenFind

Same lab was screened as EXCLUDED under the earlier EDBB pass — flagged as a reconciliation point.

GANs / latent-space methods for generative architectural design

"Design Brain" (Seoul Biennale) — DL-generated Alpine architecture

Historical NLP/OCR, ML-based 4D urban reconstruction, historical-map CV

dhSegment; Venice Time Machine; ICDAR 2026 HIPE-OCRepair

ML for urban-form indicators & planning decision-support

"Reviewing ML methods to model urban form indicators..." (2021)

Explainable AI for building-energy benchmarking

"Bench2X" two-stage XAI framework (2026)

Spatiotemporal ML for building-stock demolition/reuse risk

"Spatiotemporal machine learning for building demolition risk modelling" (2026)

Computer-vision-assisted, AI-informed timber fabrication

"Augmented Carpentry" (2025)

ML-augmented urban metabolism/emissions analysis; CV object-tracking of public space

"ML augmented global analysis of urban emission trends 2010–2022"

Conflicting screen: excluded under the EDCE pass (core pillars = Bayesian Networks/ABM, not ML) but confirmed under the EDAR pass — treat as borderline either way.

"Computational museology" — ML/CV/visual analytics for cultural heritage

CROSSINGS project (knowledge-graph/ML modeling of intangible heritage)

CNN-based ancient-glyph decipherment; computational social science

"How to Tell Ancient Signs Apart? Recognizing Maya Glyphs" (2018, MAAYA project)

Computational modeling of musical structure & style

"Distant Listening – Transitions of Tonality" (SNSF); "From Bach to the Beatles" (VW Foundation)

Computational / digital history, distant reading, algorithmic trail analysis

"Querying the Digital Archive of Science"

ML / neural nets for portfolio pricing, risk management & volatility prediction

"A machine learning approach to portfolio pricing and risk management..." (2022)

Deep learning / high-dim ML for asset pricing & stochastic discount factors

"Artificial Intelligence Asset Pricing Models", NBER WP 2025

ML applied to credit markets & fintech lending

"Predictably Unequal? The Effects of ML on Credit Markets", J. Finance 2022 (Brattle Prize)

NLP / text-embedding ML applied to patent data for innovation & IP analysis

"Text-Embedding-based Approach to Measure Patent-to-Patent Technological Similarity" (2022)

Deep neural networks for new-product sales forecasting

"ML-based framework for forecasting sales of new products...using DNNs" (2022)

ML / data mining for student modeling, knowledge tracing & fairness in ed-tech

Lab mission: algorithmic student-knowledge estimation and fairness

Human-AI complementarity; AI-driven classroom orchestration & analytics

"Designing for human-AI complementarity in K-12 education", AI Magazine 2022; Lumilo

Learning-analytics infrastructure; explainable AI on educational data

"An End-to-End Data Pipeline for Managing Learning Analytics"; GraphNEx

Context-aware generative AI / fine-tuned LLMs supporting instructors

"AI Meets Mathematics Education: Supporting an Instructor with Context-Aware AI", CHI 2026