Anticancer peptides (ACPs) are short bioactive sequences that selectively target tumor cells with minimal toxicity, positioning them as promising candidates for next-generation cancer therapies. However, existing computational models face limitations in sequence representation and class imbalance. To address these challenges, we propose UACD-ACPs, a unified fusion-driven framework that integrates a diffusion-inspired noise-conditioned classifier for ACP prediction and a diffusion-based peptide generation module with cancer-type-aware organization for targeted downstream screening. The classification module integrates ProtBERT-based semantic embeddings with physicochemical descriptors via the Multiscale Embedding Compression Strategy (MECS) and a diffusion-inspired noise-conditioned encoder, substantially enhancing predictive robustness and accuracy, particularly under challenging imbalanced multi-class settings. In the generative pipeline, we introduce a denoising diffusion-based generative framework augmented by two novel fusion modules: the Bitemporal Fusion Module (BFM) and the Temporal Feature Attention Module (TFAM). These modules perform multi-scale temporal and semantic fusion to promote the generation of structurally coherent and functionally relevant peptide candidates. Experimental results demonstrate that UACD-ACPs outperforms state-of-the-art methods in terms of accuracy, F1-score, and AUC-ROC. The generated peptides exhibit favorable physicochemical properties, diverse secondary structures, and strong structural stability, as validated by molecular dynamics simulations and membrane-binding analyses. Overall, this study highlights the potential of fusion-driven diffusion-based frameworks for alleviating class imbalance and data heterogeneity in anticancer peptide modeling, paving the way for scalable and biologically grounded ACP discovery.

Therapeutic peptides play a key role in regulating cellular functions and repairing damaged cells through targeted molecular interactions. Traditional wet-lab methods for identifying therapeutic peptides rely on time-consuming biochemical assays and low-throughput screening techniques, which struggle to capture complex sequence-stability relationships critical for drug development. To address these limitations, we innovatively integrate a pretrained protein language model with stacked bidirectional long short-term memory (BiLSTM) encoders. This hybrid architecture enables hierarchical extraction of both global contextual patterns (via the language model) and localized sequential dependencies (via BiLSTM), effectively modeling nonlinear correlations within peptide sequences. A key technical lies in the proposed kernelized Takagi-Sugeno-Kang fuzzy system (K-TSK-FS), which combines fuzzy logic with kernel methods to handle sequence ambiguity while maintaining interpretability. Unlike conventional classifiers, this system maps high-dimensional features into a reproducing kernel Hilbert space, enhancing discrimination between therapeutic and non-therapeutic peptides through nonlinear decision boundaries. To evaluate the model, six benchmark datasets are used to test our model. Experimental results show that our method achieves better prediction performance.

Therapeutic peptides are therapeutic agents synthesized from natural amino acids, which can be used as carriers for precisely transporting drugs and can activate the immune system for preventing and treating various diseases. However, screening therapeutic peptides using biochemical assays is expensive, time-consuming, and limited by experimental conditions and biological samples, and there may be ethical considerations in the clinical stage. In contrast, screening therapeutic peptides using machine learning and computational methods is efficient, automated, and can accurately predict potential therapeutic peptides. In this study, a k-nearest neighbor model based on multi-Laplacian and kernel risk sensitive loss was proposed, which introduces a kernel risk loss function derived from the K-local hyperplane distance nearest neighbor model as well as combining the Laplacian regularization method to predict therapeutic peptides. The findings indicated that the suggested approach achieved satisfactory results and could effectively predict therapeutic peptide sequences.

The detection of therapeutic peptides is a topic of immense interest in the biomedical field. Conventional biochemical experiment-based detection techniques are tedious and time-consuming. Computational biology has become a useful tool for improving the detection efficiency of therapeutic peptides. Most computational methods do not consider the deviation caused by noise. To improve the generalization performance of therapeutic peptide prediction methods, this work presents a sequence homology score-based deep fuzzy echo-state network with maximizing mixture correntropy (SHS-DFESN-MMC) model. Our method is compared with the existing methods on eight types of therapeutic peptide datasets. The model parameters are determined by 10 fold cross-validation on their training sets and verified by independent test sets. Across the 8 datasets, the average area under the receiver operating characteristic curve (AUC) values of SHS-DFESN-MMC are the highest on both the training (0.926) and independent sets (0.923).

Detection of therapeutic peptide is a major research direction in the current biopharmaceutical field. However, traditional biochemical experimental detection methods take a lot of time. As supplementary methods for biochemical experiments, the computational methods can improve the efficiency of therapeutic peptide detection. Currently, most machine learning-based therapeutic peptide identification algorithms do not consider the processing of noisy samples. We propose a therapeutic peptide classifier, called weighted echo state networks based on subspace projection (WESN-SP), which reduces the bias caused by high-dimensional noisy features and noisy samples. WESN-SP is trained by sparse Bayesian learning algorithm (SBL) and introduces a weight coefficient for each sample by kernel dependence maximization-based subspace projection. The experimental results show that WESN-SP has better performance than other existing methods.

The detection of therapeutic peptides is usually a biochemical experimental method, which is time-consuming and labor-intensive. Lots of computational biology methods had been proposed to solve the problem of therapeutic peptide prediction. However, the existing methods did not consider the processing of noisy samples. We propose a kernel risk-sensitive mean p-power error-based echo state network with sparse learning (KRP-ESN-SL). An efficient iterative optimization algorithm is used to train the model. The KRP-ESN-SL has better performance than other methods.

Therapeutic peptides act on the skeletal system, digestive system and blood system, have antibacterial properties and help relieve inflammation. In order to reduce the resource consumption of wet experiments for the identification of therapeutic peptides, many computational-based methods have been developed to solve the identification of therapeutic peptides. Due to the insufficiency of traditional machine learning methods in dealing with feature noise. We propose a novel therapeutic peptide identification method called Structured Sparse Regularized Takagi-Sugeno-Kang Fuzzy System on Within-Class Scatter (SSR-TSK-FS-WCS). Our method achieves good performance on multiple therapeutic peptides and UCI datasets.