Retrospective <i>in silico</i> HLA predictions from COVID-19 patients reveal alleles associated with disease prognosis
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 Human Leukocyte Antigen (HLA) gene locus plays a fundamental role in human immunity, and it is established that certain HLA alleles are disease determinants. METHODS: By combining the predictive power of multiple in silico HLA predictors, we have previously identified prevalent HLA class I and class II alleles, including DPA1*02:02, in two small cohorts at the COVID-19 pandemic onset. Since then, newer and larger patient cohorts with controls and associated demographic and clinical data have been deposited in public repositories. Here, we report on HLA-I and HLA-II alleles, along with their associated risk significance in one such cohort of 126 patients, including COVID-19 positive (n=100) and negative patients (n=26). RESULTS: We recapitulate an enrichment of DPA1*02:02 in the COVID-19 positive cohort (29%) when compared to the COVID-negative control group (Fisher's exact test [FET] p=0.0174). Having this allele, however, does not appear to put this cohort's patients at an increased risk of hospitalization. Inspection of COVID-19 disease severity outcomes reveal nominally significant risk associations with A*11:01 (FET p=0.0078), C*04:01 (FET p=0.0087) and DQA1*01:02 (FET p=0.0121). CONCLUSIONS: While enrichment of these alleles falls below statistical significance after Bonferroni correction, COVID-19 patients with the latter three alleles tend to fare worse overall. This is especially evident for patients with C*04:01, where disease prognosis measured by mechanical ventilation-free days was statistically significant after multiple hypothesis correction (Bonferroni p = 0.0023), and may hold potential clinical value.
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 | 0.001 |
| 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 | 0.001 |
| 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