Harvesting Classification Trees for Drug Discovery
Bibliographic record
Abstract
Millions of compounds are available as potential drug candidates. High throughput screening (HTS) is widely used in drug discovery to assay compounds for a particular biological activity. A common approach is to build a classification model using a smaller sample of assay data to predict the activity of unscreened compounds and hence select further compounds for assay. This improves the efficiency of the search by increasing the proportion of hits found among the assayed compounds. In many assays, the biological activity is dichotomized into a binary indicator variable; the explanatory variables are chemical descriptors capturing compound structure. A tree model is interpretable, which is key, since it is of interest to identify diverse chemical classes among the active compounds to serve as leads for drug optimization. Interpretability of a tree is often reduced, however, by the sheer size of the tree model and the number of variables and rules of the terminal nodes. We develop a "tree harvesting" algorithm to filter out redundant "junk" rules from the tree while retaining its predictive accuracy. This simplification can facilitate the process of uncovering key relations between molecular structure and activity and may clarify rules defining multiple activity mechanisms. Using data from the National Cancer Institute, we illustrate that many of the rules used to build a classification tree may be redundant. Unlike tree pruning, tree harvesting allows variables with junk rules to be removed near the top of the tree. The reduction in complexity of the terminal nodes improves the interpretability of the model. The algorithm also aims to reorganize the tree nodes associated with the interesting "active" class into larger, more coherent groups, thus facilitating identification of the mechanisms for activity.
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.
How this classification was reachedexpand
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 | 0.001 | 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 | 0.009 |
| 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".