BioMoR - Bioinformatics Modeling with Recursion and Autoencoder-Based Ensemble

Provides tools for bioinformatics modeling using recursive transformer-inspired architectures, autoencoders, random forests, XGBoost, and stacked ensemble models. Includes utilities for cross-validation, calibration, benchmarking, and threshold optimization in predictive modeling workflows.

Last updated

autoencoderbenchmarkingbioinformaticscalibrationdeep-learningensemble-learningfeaturemachine-learningmixture-of-recursions

4.65 score 5 scripts 172 downloads

annotaR - Tidy, Integrated Gene Annotation

A framework for intuitive, multi-source gene and protein annotation, with a focus on integrating functional genomics with disease and drug data for translational insights. Methods used include g:Profiler (Raudvere et al. (2019) <doi:10.1093/nar/gkz369>), biomaRt (Durinck et al. (2009) <doi:10.1038/nprot.2009.97>), and the Open Targets Platform (Koscielny et al. (2017) <doi:10.1093/nar/gkw1055>).

Last updated

bioinformatics-pipelinedisease-predictiondrug-discoverygene-annotationgene-enrichment

3.18 score 5 scripts 2 downloads

resLIK - Representation-Level Control Surfaces for Reliability Sensing

Implements the Representation-Level Control Surfaces (RLCS) paradigm for ensuring the reliability of autonomous systems and AI models. It provides three deterministic sensors: Residual Likelihood (ResLik) for population-level anomaly detection, Temporal Consistency Sensor (TCS) for drift and shock detection, and Agreement Sensor for multi-modal redundancy checks. These sensors feed into a standardized control surface that issues 'PROCEED', 'DEFER', or 'ABSTAIN' signals based on strict safety invariants, allowing systems to detect and react to out-of-distribution states, sensor failures, and environmental shifts before they propagate to decision-making layers.

Last updated

3.00 score 4 scripts 149 downloads