# Geometric Deep Learning

Late-stage functionalization in medicinal chemistry is limited by scarce reaction data, especially when several similar C–H bonds compete for borylation. A Roche and ETH team used an XGBoost-ensemble active learning loop to pick 50 previously unreported substrates from a 22,253-molecule pool across three rounds of high-throughput experiments, expanding C–H borylation data to 6,865 reactions over 568 substrates. Geometric graph neural networks trained on that set beat XGBoost in most cases on random, Butina and Bemis–Murcko scaffold splits, and in a prospective test on ten unseen substrates with N-heteroaryl motifs the top predicted site matched experiment in all cases.

Late-stage functionalization in medicinal chemistry is limited by scarce reaction data, especially when several similar C–H bonds compete for borylation. A Roche and ETH team used an XGBoost-ensemble active learning loop to pick 50 previously unreported substrates from a 22,253-molecule pool across three rounds of high-throughput experiments, expanding C–H borylation data to 6,865 reactions over 568 substrates. Geometric graph neural networks trained on that set beat XGBoost in most cases on random, Butina and Bemis–Murcko scaffold splits, and in a prospective test on ten unseen substrates with N-heteroaryl motifs the top predicted site matched experiment in all cases.

Spatial multi-omic data pose two jobs at once: merging molecular layers such as RNA and protein measured on the same tissue section, and transferring cell-type labels from a reference dataset onto a new sample. DIRAC treats both as domain adaptation, using a graph neural network and an adversarial discriminator to learn embeddings that are insensitive to batch and modality but discriminative for cell type. On simulated and real DBiT-seq data it beat comparison methods on domain recovery and mixing, and in glioblastoma tissue it identified a microvascular proliferation domain the other methods missed.

Spatial multi-omic data pose two jobs at once: merging molecular layers such as RNA and protein measured on the same tissue section, and transferring cell-type labels from a reference dataset onto a new sample. DIRAC treats both as domain adaptation, using a graph neural network and an adversarial discriminator to learn embeddings that are insensitive to batch and modality but discriminative for cell type. On simulated and real DBiT-seq data it beat comparison methods on domain recovery and mixing, and in glioblastoma tissue it identified a microvascular proliferation domain the other methods missed.

SimCortex v2 reconstructs the left and right white-matter and pial cortical surfaces jointly in one shared deformation space, matching the strongest baseline's surface-distance accuracy on 560 T1-weighted MRI cases from 14 cohorts while showing no detected inter-surface collision in 92.14% of cases, whereas every baseline produced at least one collision in every case. It initializes topologically correct meshes from a volumetric segmentation, then predicts multi-scale stationary velocity fields with a ribbon-conditioned U-Net and regularizes surface separation during training.

SimCortex v2 reconstructs the left and right white-matter and pial cortical surfaces jointly in one shared deformation space, matching the strongest baseline's surface-distance accuracy on 560 T1-weighted MRI cases from 14 cohorts while showing no detected inter-surface collision in 92.14% of cases, whereas every baseline produced at least one collision in every case. It initializes topologically correct meshes from a volumetric segmentation, then predicts multi-scale stationary velocity fields with a ribbon-conditioned U-Net and regularizes surface separation during training.

This work applies message-passing graph neural networks to Inverse Beta Decay detection of the Diffuse Supernova Neutrino Background in water-Cherenkov detectors, building a pipeline that reconstructs low-energy events from raw photomultiplier hits: one model regresses 3–30 MeV positron energy and another classifies neutron-capture signals, achieving performance comparable to the state of the art on simulations in a Hyper-Kamiokande-like geometry.

This work applies message-passing graph neural networks to Inverse Beta Decay detection of the Diffuse Supernova Neutrino Background in water-Cherenkov detectors, building a pipeline that reconstructs low-energy events from raw photomultiplier hits: one model regresses 3–30 MeV positron energy and another classifies neutron-capture signals, achieving performance comparable to the state of the art on simulations in a Hyper-Kamiokande-like geometry.