2026-10-14T17:48:02Z https://api.figshare.com/v2/oai

oai:figshare.com:article/34033789 2026-09-30T16:09:48Z category_1 category_4 category_7 category_13 category_15 category_21 category_45 category_734 category_122 portal_63 item_type_6 month_year_09_2026

Fréchet ChemNet Distance Is Blind to Double-Bond Stereochemistry Jinseong Yim (25144828) Biophysics Biochemistry Medicine Genetics Neuroscience Biotechnology Sociology Biological Sciences not elsewhere classified Mental Health two models  chemnet tokenizer maps benchmark data contain 384 molecules altered guacamol cannot expose bond accuracy falls inverting every double 26 versus 0 cannot distinguish smiles bond bond configuration z </ unpaired generated unknown token statistically borderline reported directly molecular skeleton metric operating fold difference e </ double bonds distributional score direction symbols dependable signal crosses zero conditional per caption correctness bootstrap interval aware correction 20 output 03 %, Fréchet ChemNet Distance (FCD) cannot distinguish <i>E</i>/<i>Z</i> stereoisomers, because the ChemNet tokenizer maps the SMILES bond-direction symbols / and <i>\</i> to the same unknown token; inverting every double-bond configuration in MolT5-large’s ChEBI-20 output (384 molecules altered) changes FCD by 0.001 resampling standard deviations while conditional per-bond accuracy falls from 98.97% to 1.03%, and MOSES and GuacaMol cannot expose the defect because their benchmark data contain no stereochemistry. A stereochemistry-aware correction is monotone under synthetic corruption yet prefers fully inverted real output, illustrating that a metric operating only on unpaired generated and reference embeddings cannot recover per-caption correctness. Across real generators FCD carries no dependable signal about a more-than-3-fold difference in <i>E</i>/<i>Z</i> recall: on the double bonds where MolT5-large and BioT5-base both recover the molecular skeleton, recall is 0.26 versus 0.997, yet FCD does not separate the two modelsits ordering is statistically borderline, with a bootstrap interval that crosses zero. The failure is one of uninformativeness rather than of reversal. Where paired conditional references exist, stereochemistry should be reported directly, as coverage, precision, and recall, rather than inferred from a distributional score. 2026-09-30T00:00:00Z Text Journal contribution 10.1021/acs.jcim.6c02441.s001 https://figshare.com/articles/journal_contribution/Fre_chet_ChemNet_Distance_Is_Blind_to_Double-Bond_Stereochemistry/34033789 CC BY-NC 4.0