Primer set evolution

Live view of the evolutionary optimizer: best set coverage is the union of sequences hit by the current best individual (max over the population), relative to 1,110,935 Layer‑3 sequences. The worker updates Redis each generation; this page uses SSE for near real-time updates.

Run evolution

Queues a Celery job on the cluster. You can only have one run at a time while status is running. If you use Rebuild pair cache, expect roughly 30–120+ minutes of RocksDB I/O before the GA starts; the dashboard should show phases load_kmers and pairs during that time.

Without the checkbox, the job reuses /data/evolution/pair_csr.npz if it exists. First deployment needs either a checked run or a manual cache build on the server.

Best set coverage (current evolution)
Sequences covered: —
0% of coverable index
idle

Run metrics

Mean population coverage
Generation
Phase
Elapsed
Population
Pairs per individual

Best coverage over time (this session)

Event log