Subtyping Schizophrenia Using Psychiatric Polygenic Scores
Bibliographic record
Abstract
Abstract Background Subtyping schizophrenia can disentangle heterogeneity and help with treatment decision- making. However, current schizophrenia subtypes have not demonstrated adequate clinical utility, limited by sample size, suboptimal clustering methods, and choice of clustering input. Polygenic scores (PGS) reflect the genetic risk of phenotypes including comorbidities and are available before treatment, making them candidate clustering input. Methods We derived PGS for schizophrenia, autism spectrum disorder, bipolar disorder type-1, depression, and intelligence in 4,915 schizophrenia cases with register linkage. We randomly divided the sample into discovery and replication partitions and applied a novel clustering workflow on both: preprocessing PGS, feature extraction with uniform manifold approximation and projection (UMAP), and clustering with density-based spatial clustering of applications with noise (DBSCAN). After replication, we re-performed clustering on the entire sample and evaluated treatment-relevant variables of medication and hospitalization (extracted from registers) across clusters. Outcomes We identified five well-replicated PGS clusters. Cluster 1 (26% of entire sample) with generally lower PGS, had the least use of antipsychotics (including clozapine), and fewer outpatient visits. Cluster 2 (48%) with generally higher PGS, especially schizophrenia PGS, had more prescriptions of antipsychotics including clozapine and longer treatment with clozapine. Each featured by specific PGS, clusters 3 (high IQ-PGS, 11%), 4 (high ASD-PGS, 8%), 5 (high BIP-PGS, 7%) showed sub-threshold level significance in the corresponding phenotypic measures but did not differ significantly in the treatment-relevant variables. Solely categorizing the patients with SCZ-PGS did not generate any significant patterns in the phenotypic and treatment-relevant variables. Interpretation The results suggest that combinations of PGS of brain disorders and traits can provide clinically relevant clusters, offering a direction for future research on schizophrenia subtyping. Future replications in independent samples are required. The workflow can be generalized to other disorders and with mechanism-informed PGS.
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 | 0.003 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.003 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 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".