supp_tokenizer_bench / results /js_vs_random.log
nancyH's picture
Add Supp. Fig. 2b JS-distance analysis: code, region/background BEDs, FASTAs, results, baseline tokenizers
89dec6b verified
Raw History Blame Contribute Delete
1.25 kB
Tokenizing with DNAbert2...
promoter: 12,334,449 tokens
enhancer: 81,613,472 tokens
exon: 39,014,803 tokens
random: 32,378,182 tokens
Tokenizing with Grover...
promoter: 15,845,529 tokens
enhancer: 102,460,422 tokens
exon: 49,411,407 tokens
random: 41,049,736 tokens
Tokenizing with cCRE_region_BPE...
promoter: 12,016,320 tokens
enhancer: 77,428,659 tokens
exon: 37,654,355 tokens
random: 31,601,798 tokens
Tokenizing with motif_region_BPE...
promoter: 13,091,114 tokens
enhancer: 83,353,540 tokens
exon: 40,627,217 tokens
random: 33,666,692 tokens
Tokenizing with DNAMotifTokenizer...
promoter: 19,543,735 tokens
enhancer: 126,777,588 tokens
exon: 61,613,820 tokens
random: 52,419,078 tokens
Saved matrix: /data/n5huang/dna_token/enrichment_analysis/js_vs_random_matrix.csv
promoter enhancer exon
DNAbert2 0.2700 0.1530 0.1658
Grover 0.2465 0.1292 0.1440
cCRE_region_BPE 0.2677 0.1527 0.1705
motif_region_BPE 0.2575 0.1444 0.1626
DNAMotifTokenizer 0.2047 0.1049 0.1227
Saved: /data/n5huang/dna_token/enrichment_analysis/js_vs_random_heatmap.pdf
Saved: /data/n5huang/dna_token/enrichment_analysis/js_vs_random_heatmap.png