Operational dashboard

The instrument

Every data, pipeline, results and operations surface behind the argument, grouped the way the pipeline runs. The book stays at the top level; this is the tab you open to operate it.

Pipeline

From sequence to GO term

EmbeddingsPLM embedding configs across the eight backbones
KNN SearchEmbedding-similarity GO annotation, BPO / MFO / CCO
ScoringCombine distance, alignment, taxonomy, evidence
Re-rankerLightGBM reranker over scored predictions
DatasetsFrozen reranker dumps and export dispatcher

Reference Data

Reference proteome, ground truth, query sets

ProteinsUniProt entries, Swiss-Prot + TrEMBL, isoforms
GO AnnotationsGO ontology snapshots and ground-truth sets
Query SetsFASTA uploads grouped for batch runs

Results

Benchmark matrix and CAFA evaluation

BenchmarkThe nine-cell benchmark matrix
Experiment graphNodes, edge strength and the nine panels, read off the record
CAFA EvaluationCAFA-style delta evaluation

Operations

Queues, maintenance, stack health

JobsLive job queue and event audit trail
StackEight repositories, open PRs, deploy targets
Agent farmAgent-farm tasks, plan DAG and cost