The study aims to develop an in silico method for predicting the toxicity of peptides and proteins, which is crucial for the development of peptide/protein-based therapies. The researchers collected toxic peptides with 35 or fewer residues from various databases and non-toxic peptides from SwissProt and TrEMBL. They observed that certain residues like Cys, His, Asn, and Pro are abundant in toxic peptides. Machine learning techniques, specifically support vector machines (SVM), were used to develop models for predicting toxicity based on amino acid and dipeptide compositions. The dipeptide-based model achieved an accuracy of 94.50% with a Matthew's correlation coefficient (MCC) of 0.88. A hybrid model combining motif information with the dipeptide-based SVM further improved the performance to 98.41%. The researchers also developed a web server called ToxinPred, which can predict toxicity, identify toxic regions in proteins, and design peptides with desired toxicity. The study highlights the importance of in silico methods in saving time and resources in drug discovery and provides a valuable tool for the scientific community.The study aims to develop an in silico method for predicting the toxicity of peptides and proteins, which is crucial for the development of peptide/protein-based therapies. The researchers collected toxic peptides with 35 or fewer residues from various databases and non-toxic peptides from SwissProt and TrEMBL. They observed that certain residues like Cys, His, Asn, and Pro are abundant in toxic peptides. Machine learning techniques, specifically support vector machines (SVM), were used to develop models for predicting toxicity based on amino acid and dipeptide compositions. The dipeptide-based model achieved an accuracy of 94.50% with a Matthew's correlation coefficient (MCC) of 0.88. A hybrid model combining motif information with the dipeptide-based SVM further improved the performance to 98.41%. The researchers also developed a web server called ToxinPred, which can predict toxicity, identify toxic regions in proteins, and design peptides with desired toxicity. The study highlights the importance of in silico methods in saving time and resources in drug discovery and provides a valuable tool for the scientific community.