ApexOracle paper assets: pathogen embeddings, MIC candidate scorer, and classification reproduction data

This version extends the paper-listed ApexOracle data release while retaining the original precomputed genome and strain-text embedding archives. Genome_embs.tar.gz: precomputed Evo-2 genome embeddings. Text_Description.tar.gz: strain descriptions and precomputed text embeddings. mic_candidate_scorer_all_peptide_non_pad_t1e-3_epoch13.pth: the all-peptide post-generation MIC candidate-scoring checkpoint. apexoracle_fig1b_classification_reproduction_v1.tar.gz: exact Fig. 1b folds, normalized sample-level predictions, and a standard-library metric checker. The MIC scorer uses the fixed_epsilon_non_pad profile at t=1e-3; it is not an exact clean t=0 model, is not the random-time padding-preserved Generation sampler checkpoint, and should not be represented as a validated general small-molecule scorer. See zenodo_release_21882300.json for per-file roles, sizes, MD5, SHA-256, licenses, and scientific boundaries.