turns out this matters a lot for ML , hypergraphs capture way more context and generalize better
made some slides if you're curious about the difference and why it matters for learning networks

Welcome, if you want to come to this place, it is fun. Search for it (Holy Karbala), the burial place of the Master of Martyrs, Al-Hussein bin Ali bin Abi Talib.

Iraq\Karbala

And this isn’t a one-shot: results come from 3 model configs × 2 seed partitions → stable + robust.
Even vs. papers like DeepDDI, NMDADNN, SumGNN, HetDDI (with heavy KGs + multi-modal data), this SMILES-only design holds up.

Next: improve rare classes, but message is clear → chemistry drives metabolism

And this isn’t a one-shot: results come from 3 model configs × 2 seed partitions → stable + robust.
Even vs. papers like DeepDDI, NMDADNN, SumGNN, HetDDI (with heavy KGs + multi-modal data), this SMILES-only design holds up.

classes, but message is clear → chemistry drives metabolism.

And this isn’t a one-shot: results come from 3 model configs × 2 seed partitions → stable + robust.
Even vs. papers like DeepDDI, NMDADNN, Next: improve rare
SumGNN, HetDDI (with heavy KGs + multi-modal data), this SMILES-only design holds up.

classes, but message is clear → chemistry drives metabolism.

And this isn’t a one-shot: results come from 3 model configs × 2 seed partitions → stable + robust.
Even vs. papers like DeepDDI, NMDADNN, Next: improve rare
SumGNN, HetDDI (with heavy KGs + multi-modal data), this SMILES-only design holds up.

Beyond strong benchmarks, we extended the model to multi-task learning: not only detecting interactions but also classifying their type ; especially metabolic ones vital for clinicians.