Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-0211
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-07-13T22:15:28.795220Z
status: published
lens: g_space
sector: science
headline: Accurate prediction of protein structures and interactions using a three-track neural network
dek: Deep learning takes on protein folding In 1972, Anfinsen won a Nobel prize for demonstrating a connection between a protein’s amino acid sequence and its three-dimensional structure. Since 1994, scientists have competed in the biannual Critical Assessment of Structure Prediction (CASP) protein-folding challenge. Deep learning methods took center stage at CASP14, with DeepMind’s Alphafold2 achieving remarkable accuracy. Baek et al . explored network architectures based on the DeepMind framework. They used a three…
gain_title: A three-track neural network that jointly processes sequence, distance, and coordinate information enables accurate prediction of protein structures and protein-protein complexes, approaching DeepMind's accuracy.
problem_title: (none)
trace_subject: (none)
gain_reading: A three-track neural network that jointly processes sequence, distance, and coordinate information enables accurate prediction of protein structures and protein-protein complexes, approaching DeepMind's accuracy.
gain_evidence: They used a three-track network to process sequence, distance, and coordinate information simultaneously and achieved accuracies approaching those of DeepMind.
problem_reading: (none)
problem_evidence: (none)
quick_read: In work published August 20, 2021, researchers built on the CASP14-era DeepMind approach to protein folding. They developed RoseTTAFold, a three-track network that processes sequence, distance, and coordinate information simultaneously, achieving accuracies approaching those of DeepMind.

The advance matters because it provides a more accessible route to accurate structure determination for experimental biology, potentially accelerating work that depends on x-ray crystallography, cryo-electron microscopy, and protein-complex modeling. What remains uncertain from this summary is how broadly the accuracy holds across protein classes and how performance compares on independent benchmarks.
limitation: 
tag: Evidence-backed gain
key_points: Method named RoseTTAFold builds on DeepMind framework presented at CASP14. | Architecture processes three information tracks simultaneously: sequence, distance, and coordinates. | Demonstrated utility for solving x-ray crystallography and cryo-electron microscopy modeling problems.
rundown: Baek et al. explored network architectures based on the DeepMind framework that featured at CASP14, where DeepMind's AlphaFold2 achieved remarkable accuracy.

Their RoseTTAFold implementation was reported to solve challenging experimental modeling tasks and to generate accurate models of protein-protein complexes, linking amino acid sequence to three-dimensional structure.
sources:
- peer_reviewed | Science | https://doi.org/10.1126/science.abj8754 | 2021-08-20
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

e51b82b8827f8b185ee6edb2c78ba30826a86519c8b1d024ffa4e679742e63cd

previous

0000000000000000000000000000000000000000000000000000000000000000