blast_samples/ ================ Sample sequences prepared for Aditya's bidirectional (reciprocal best hit) BLAST validation of haploblock cluster consensus sequences, as requested by Ben Busby for the Copenhagen hackathon collaboration. Source locus: chr1:10023878-10175888 FILES ----- consensus_samples.fasta 9 representative sequences: 3 BIG (cluster size >=1000), 3 SMALL (5-15), 3 SINGLETON (size 1), drawn from three regions -- chr1:10023878-10175888, chr2:100118009-100180909, chr3:100195058-100330759. Headers are labeled CATEGORY_block_sizeN_originalID. full_big_cluster.fasta Complete membership (2428 sequences) of the chr1 BIG cluster (representative HG00410_chr1_region_10023878-10175888_hap1). full_big_cluster_members.tsv Raw clusters_summary.tsv rows (chr, block, start, end, cluster_id, member, hap, cluster_size) for every sequence in full_big_cluster.fasta. Provenance/audit trail, not sequence data. full_small_cluster.fasta Complete membership (5 sequences) of the chr1 SMALL cluster (representative HG00406_chr1_region_10023878-10175888_hap1). full_small_cluster_members.tsv Same as above, for the SMALL cluster. representative_candidates.tsv Selection table for all 9 consensus picks (category, chr, block, start, end, cluster_id, member, hap, cluster_size). See blast_samples_chr2_chr3/ and blast_samples_extended/ for follow-up tests of generalizability and size-dependence. GENERATION ---------- Produced by select_blast_samples.sh from clusters_summary.tsv and /clusters/_{rep_seq,all_seqs}.fasta. Full-cluster FASTA files re-extracted with reextract_with_fixed_dedup.sh on 2026-09-05 to correct the duplicate-representative bug described above.