For developing computational methods to predict the effects of cancer mutations and for evaluating the false positive rates of cancer driver mutation detection efforts, benchmark datasets containing both positive (driver mutations) and negative (passenger mutations) datasets are needed. While some recently developed datasets for cancer driver mutations are available for this purpose, benchmark sets for cancer passenger mutations have largely been missing. Therefore, we developed a comprehensive literature-based database of Cancer Passenger Mutations (dbCPM). In the current release, dbCPM contains 941 experimentally supported and 978 putative passenger mutations derived by manual curation of literature. All mutation descriptions include the position information (gDNA, cDNA, and protein levels), specific diseases, evidence sentences as well as evidence levels (in vivo, in vitro, or putative). The database can facilitate the development and evaluation of computational methods for predicting the effects of cancer mutations.

To obtain genomic positions, we used TransVar (http://bioinformatics.mdanderson.org/transvarweb/) through entering genes and their associated cDNA changes and mapped them to the results of the longest possible transcripts. Genomic positions that failed to match the canonical cDNA at the specified site were substituted by a dot (.). To acquire standard disease terminology, we mapped the related disease onto Disease Ontology (http://www.disease-ontology.org/).

Please cite the paper, if you are using the information in the database:

Yue Z, Zhao L, Xia J. dbCPM: a manually curated database for exploring the cancer passenger mutations. Briefings in bioinformatics, 2018. doi.org/10.1093/bib/bby105.

The dbCPM requires a modern web browser with JavaScript and cookies enabled. To view the complex details, pop-ups must not be blocked. The following browsers have been thoroughly tested with dbCPM:

• Mozilla Firefox, version 4 or above
• Internet Explorer, versions 9 or above
• Chrome, version 5 or above

The latest version of Firefox and Chrome is recommended for visualization.

Levels of evidence for mutations in dbCPM.

Level No.	Details
Level 1 (in vivo)	Alteration is regarded as a passenger based on evidence from functional experiments in vivo.
Level 2 (In vitro)	Alteration is regarded as a passenger based on evidence from functional experiments In vitro.
Level 3 (Putative)	Alteration is a putative passenger based on evidence such as a high recurrence frequency in healthy controls, an unimportant location in protein and so on.

In dbCPM, we tried to make it powerful and convenient to be used. This Usage is prepared for the online service. The dbCPM provides the browse function, search function and download function at present.

Capitalised titles correspond to column headings in the web page tables:

• DISEASE: disease terminology in Disease Ontology
• DOID: Disease Ontology identifier
• GENE: The official gene symbol
• Transcript: Transcript stabel ID from Ensembl Genome Browser
• TYPE: missense, synonymous, nonsense, indel (insertion or deletion) or noncoding
• HIGHEST LEVEL: The highest evidence level of variants across specified both disease and gene

(A) Database statistics

Gene	Mutation	Disease	PMID	Entry	Level 1 (in vivo)	Level 2 (in vitro)	Level 3 (putative)
161	1919	43	312	3813	157	1112	2544

(B) Type distribution

Missense	Synonymous	Nonsense	Indel	Noncoding	Total
1673	127	6	24	89	1919

The version 1.1 was released on June 8th, 2018.

The version 1.0 was released on January 4th, 2018.