Reproducibility crisis in ML science: Only 7 of 32 PPI methods re-trainable
anshulkundaje · x · 2026-08-17
A survey of Protein-Protein Interaction (PPI) prediction methods highlights a severe reproducibility crisis in ML-based science:
- Low code availability: Only 12 out of 32 PPI prediction methods had source code available.
- High barrier to re-training: Among those, only 7 could be re-trained with reasonable effort.
- Benchmarking: The study compared these methods against 8 baseline ML models.
More from Research
- Sentient Hosts Open AGI Builder's Day at KDD 2026 in Jeju — 0xsachi · 2026-08-17
- VoiceChat-TTS: Low-Latency Streamable TTS Supporting Mid-Utterance Barge-In — rdesh26 · 2026-08-17
- Reviewer Laments Three Shallow LLM-Generated Papers in Days — davidmanheim · 2026-08-17
- King's College London uses AI to make string theory testable for the first time — skdh · 2026-08-17
- HEP-TH arXiv Submissions Surge 30% as AI Tools Accelerate Physics Research — skdh · 2026-08-17
- Stanford CS231N 2025 taught by Fei-Fei Li is now free on YouTube — Aiden_Tech_Ai · 2026-08-17