Detection of Cancer before the physical appearance of any symptom, by definition is an arduous task to perform but due to advancement of oncology to phosphopetide level analysis the thought has been put in practice by creating a mapping and pipelines that can identify changes occurring at molecular level at initial stages of cancer initiation in body which could lead to cancer down the lane in future. The paper represents a study which is capable to identification of proteomics patterns if occurring in human body will eventually lead to cancer. The dataset used in the study includes 100 peptides across 40 samples where the anomalies have been deliberately introduced to replicate the molecular changes present in original dataset which is around 300 GB of data. The breakdown of process is in the form of data imputation, then normalizing the data before moving on to training and testing stage. The training was then performed using Random Forest classifier with which greater accuracy was obtained approximately (~95–97%) in identification of cancerous and non-cancerous peptides patterns. Analysis of important characteristic patterns of peptides has provided with varid degree of variations due to which identifications accuracy can be very high. The method so generated has been implemented in Google Collab which provide basic tool for diagnosis using proteomics data for cancer detection.