THP-NanoTarget: leakage-aware cheminformatics for tumor-homing peptide evaluation and conservative analog prioritization

Abstract Background Public tumor-homing peptide (THP) resources have enabled predictors such as THPep, SCMTHP, StackTHPred, StackTHP and LLM4THP. At the same time, broader peptide machine-learning resources, including the recent PepBenchmark/PepBenchData initiative, are moving the field toward standardized datasets, preprocessing, and leaderboards across peptide-property tasks. These developments have improved comparability but have not removed task-specific problems in THP modelling: available THP benchmarks are motif-rich, chemically redundant, and susceptible to optimistic evaluation under near-neighbor overlap and potential leakage. A cheminformatics evaluation of THP predictions therefore needs to ask not only whether a peptide receives a high score, but also whether that score is independent, motif-driven, chemically redundant, hard-negative-sensitive, or near-duplicate-dependent. Results We present THP-NanoTarget, a reproducible peptide-cheminformatics framework for leakage-aware THP benchmarking, motif-rule comparisons, Morgan/ECFP chemical-space auditing, hard-negative evaluations, strict external-recognition filtering, and conservative analog prioritization. The primary benchmark contained 4202 peptides: 1401 THP positives and 2801 primary negatives. Under standard stratified out-of-fold evaluations, a term frequency-inverse document frequency (TF-IDF) 2–4-mer logistic-regression model achieved an area under the receiver operating characteristic curve (AUROC) of 0.784 and an area under the precision recall curve (AUPRC) of 0.716, improving over a transparent motif-rule baseline (AUROC 0.644; AUPRC 0.488). Cluster-aware validation was more conservative, with an AUROC of 0.708 and an AUPRC of 0.597 for the reference k-mer model and an AUROC of 0.715 and AUPRC of 0.612 for the best rank-averaged ensemble. Chemical-neighbor exclusion revealed local-neighborhood dependence: the AUROC declined to 0.721 at a Tanimoto of < 0.80 and 0.623 at < 0.70, while the 0.60 threshold was not evaluable. A benchmark redundancy analysis showed a median nearest-neighbor Morgan/extended connectivity fingerprint (ECFP) Tanimoto similarity of 0.837, and 95.8% of peptides had at least one neighbor at ≥ 0.70. A strict external-recognition analysis retained only two of 29 literature-derived candidates after nearest-primary Tanimoto < 0.90 filtering. Conservative analog generation produced 9477 candidates, dominated by close analogs, near-duplicates, and duplicate-like variants of known positives. Conclusions THP-NanoTarget supports transparent computational triage and benchmark auditing, not de novo biological discovery or claims of receptor binding, tumor delivery, therapeutic efficacy, safety, pharmacokinetics, or clinical actionability. Scientific contribution THP-NanoTarget advances tumor-homing peptide modelling by reframing THP prediction as a leakage-aware benchmark-reliability problem rather than a conventional sequence-classification task. Unlike prior THP predictors that mainly report classification performance, it integrates motif-family annotation, transparent motif-rule baselines, RDKit/Morgan fingerprint similarity auditing, chemical-neighbor exclusion, hard-negative testing, strict external-recognition filtering, and conservative analog-novelty classification into a reproducible peptide-cheminformatics workflow. This enables THP candidate scores to be interpreted in terms of learned sequence signal, motif recurrence, near-neighbor recovery, hard-negative ambiguity, and public-data overlap, supporting transparent computational triage without claiming experimental validation.