Improving the accuracy of protein secondary structure prediction using structural alignment
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
BACKGROUND: The accuracy of protein secondary structure prediction has steadily improved over the past 30 years. Now many secondary structure prediction methods routinely achieve an accuracy (Q3) of about 75%. We believe this accuracy could be further improved by including structure (as opposed to sequence) database comparisons as part of the prediction process. Indeed, given the large size of the Protein Data Bank (>35,000 sequences), the probability of a newly identified sequence having a structural homologue is actually quite high. RESULTS: We have developed a method that performs structure-based sequence alignments as part of the secondary structure prediction process. By mapping the structure of a known homologue (sequence ID >25%) onto the query protein's sequence, it is possible to predict at least a portion of that query protein's secondary structure. By integrating this structural alignment approach with conventional (sequence-based) secondary structure methods and then combining it with a "jury-of-experts" system to generate a consensus result, it is possible to attain very high prediction accuracy. Using a sequence-unique test set of 1644 proteins from EVA, this new method achieves an average Q3 score of 81.3%. Extensive testing indicates this is approximately 4-5% better than any other method currently available. Assessments using non sequence-unique test sets (typical of those used in proteome annotation or structural genomics) indicate that this new method can achieve a Q3 score approaching 88%. CONCLUSION: By using both sequence and structure databases and by exploiting the latest techniques in machine learning it is possible to routinely predict protein secondary structure with an accuracy well above 80%. A program and web server, called PROTEUS, that performs these secondary structure predictions is accessible at http://wishart.biology.ualberta.ca/proteus. For high throughput or batch sequence analyses, the PROTEUS programs, databases (and server) can be downloaded and run locally.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it