OncoSig is a supervised Machine Learning algorithm for constructing molecular-interaction Signaling Maps (SigMaps) for an oncoprotein specific for a given tumor type. OncoSig integrates features from PrePPI and protein-protein interactions (PPIs) inferred from genomics data from, for example, patient samples with lung adenocarcinoma, by reverse engineering algorithms. A SigMap for KRAS recapitulated published KRAS biology and identified novel proteins synthetic lethal with mutant KRAS, 18 of 21 of which were validated in 3D spheroid models for LUAD. The KRAS LUAD SigMap consists of established and novel K-Ras pathway members and is enriched in known targets of FDA approved drugs. In this example script, we create a SigMap for the oncoprotein KRAS using a reduced-size version of a network file generated by processing lung adenocarcinoma (LUAD) samples, from the Genome Cancer Atlas project (TCGA). The companion Docker container includes the full version of the network, as well as a network generated from the TCGA colon adenocarcinoma (COAD) sample collection. In addition, information is provided for creating SigMaps for nine other oncoproteins: CDKN2A, EGFR, MAPK, NTRK3, PI3K, TP53, STK11, YAP1 and CTNNB1. Details about how the networks are generated can be found in the OncoSig publication.