Modified Decision Tree with Custom Splitting Logic Improves Generalization across Multiple Brains’ Proteomic Data Sets of Alzheimer’s Disease
Bibliographic record
Abstract
Many factors negatively affect a generalization of the findings in discovery proteomics. They include differentiation between patient cohorts, a variety of experimental conditions, etc. We presented a machine-learning-based workflow for proteomics data analysis, aiming at improving generalizability across multiple data sets. In particular, we customized the decision tree model by introducing a new parameter, min_groups_leaf, which regulates the presence of the samples from each data set inside the model's leaves. Further, we analyzed a trend for the feature importance's curve as a function of the novel parameter for feature selection to a list of proteins with significantly improved generalization. The developed workflow was tested using five proteomic data sets obtained for post-mortem human brain samples of Alzheimer's disease. The data sets consisted of 535 LC-MS/MS acquisition files. The results were obtained for two different pipelines of data processing: (1) MS1-only processing based on DirectMS1 search engine and (2) a standard MS/MS-based one. Using the developed workflow, we found seven proteins with expression patterns that were unique for asymptomatic Alzheimer patients. Two of them, Serotransferrin TRFE and DNA repair nuclease APEX1, may be potentially important for explaining the lack of dementia in patients with the presence of neuritic plaques and neurofibrillary tangles.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.005 | 0.010 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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 source (direct Gemma or distilled Codex), 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".