Key (lock)×Computational biology

As computational biology industrializes around large models (AlphaFold, PPI networks, phylogenomics), the raw material — genomic, clinical, and wearable-derived biological data — becomes a high-value asset that must be locked, keyed, and selectively unlocked. The collision zone is cryptographic and access-control machinery (keys/locks) applied natively to biological computation: privacy-preserving inference, encrypted genomic search, and biometric/physiological key generation.

The two fields share unusually functional bridges — Data science, Information retrieval, Wearable technology, and Process (computing) — meaning the plumbing (encrypted search, streaming physiological signals, pipeline orchestration) already spans both. With 34 common neighbours and an Adamic-Adar affinity of 8.36, plus 13 authors already publishing on both sides separately, the human and methodological substrate exists even though direct co-publication does not. Field A's high recent-share (0.134) signals a community accelerating faster than the larger, more mature computational-biology base.

Teams fluent in both privacy-preserving cryptography (homomorphic encryption, federated learning, secure multiparty computation) and modern bio-ML pipelines will win — likely emerging from health-data institutions and genomics platforms that already own sensitive data and are forced to compute on it without exposing it. The interpretability strand in Field A ('use interpretable models for high-stakes decisions') suggests the winners will also pair access control with auditable, explainable models for regulated clinical use.

A benchmark and reference implementation for privacy-preserving protein/variant search: run AlphaFold-style structure or PPI-network queries against an encrypted genomic database using homomorphic encryption or secure enclaves, measuring accuracy loss vs. plaintext and latency at cohort scale — with interpretable, auditable outputs for clinical decision-making.