• User's recommendations
• Principle and system requirement
• Mutation pattern
• Mutation effect
• Codon distribution
• Function pattern
• Tumor spectrum
• Advanced search
• Cell-line search
• Function analyses
• Mutation prevalence
• Mutation validation tool
• Structure analysis
• Data retrieval and download

• Somatic dataset
• Germline dataset
• Prevalence dataset
• Prognosis dataset
• Function datasets
• Cell-lines
• Gene variations
• Polymorphisms
• Mouse models

• Mutation
• Tumor site and type
• Demographic Information
• Reference
• Mutation detection method

• A...E
• F...J
• K...O
• P...T
• U...Z

The database is exclusively based on peer-reviewed literature reports and represents a repository for all published TP53 mutation data. Users should be aware of the following limitations and possible biases that may affect the analysis of the database:

Publication of the same set of mutations in different papers by the same authors is a serious problem that has led to duplicates entries in the database in the past. We now perform systematic searches of the database under the author’s name to identify earlier entries that may correspond to the same mutations. We have also extensively reviewed the entire dataset in order to find and eliminate these duplications. However, despite these efforts, some duplicates may remain in the database and their identification is an ongoing task.
Another problem arises from the fact that publications may contain data that are either erroneous or affected by experimental biases. Our policy is to enter all published data into the database, with the idea that the database should represent a repository for all the published TP53 mutation data. It is the responsibility of database users to make their own critical judgement on the validity of the data that they include in their analyses. Indeed, it is possible that sets of data that seem aberrant will later appear to correspond to very rare mutations occurring only in specific population groups.
Editorial policies have a major impact on the reporting of TP53 mutations. In particular, studies that confirm earlier findings in common cancers are often considered as repetitive and thus rejected. Moreover, many recent studies do not provide detailed information on each mutation detected but rather report their results in the form of summary tables or graphs. This information can't be entered in the database, resulting in the loss of important data. Thus, database users should be aware that the dataset cannot be considered as an exhaustive description of the frequency, nature and distribution of TP53 mutations in human cancers.

Most of the data included in the database are from retrospective studies where tumor cases are selected on the basis of relevance to pathological questions. Less than 10% of publications are molecular epidemiological studies with information on individual exposures. The lack of detailed information on exposure is a major concern for the correct interpretation of mutation patterns.
Another potential source of bias lies in the methodology used for mutation detection. The mode of tissue fixation and conservation may affect the sensitivity of the detection. Each of the commonly used, pre-screening methods (SSCP, DGGE, yeast functional assay) has its own limitations in terms of sensitivity and specificity. Many studies are limited to exons 5-8, because initial reports have shown that mutations tend to cluster in this region. Database users should also be aware that some studies might be affected by important technical artifacts. Such artifacts may have an impact on the interpretations of rare mutation patterns. 

Data annotation (on tumor pathology, individual exposures, clinical parameters and cancer outcome) involves an effort of standardization. For example, we have adopted the International Classification of Disease for Oncology as a standard for tumor sample description. As many studies use different classifications, we often have to translate these classifications to ICD-O standards. The same applies to data on individual exposure (such as for example tobacco or alcohol consumption, viral infections, etc...), and to the description of mutations. Mutations are sometimes poorly described or only described at the RNA or protein level, while we annotate them at the genomic level (more in the FAQ). In the past years, we have updated our annotation system several times in order to improve its accuracy.

We recommend reading the two following publications before using the database:

  Petitjean A, Mathe E, Kato S, Ishioka C, Tavtigian SV, Hainaut P, Olivier M. Impact of mutant p53 functional properties on TP53 mutation patterns and tumor phenotype: lessons from recent developments in the IARC TP53 database. Hum Mutat. 2007 Jun;28(6):622-9 PubMed

  Hernandez-Boussard T, Montesano R, Hainaut P. Sources of bias in the detection and reporting of p53 mutations in human cancer: analysis of the IARC p53 mutation database. Genet Anal. 1999;14(5-6):229-33. PubMed

The web site has been optimized for PC computers using Internet Explorer 5 or later versions. Searching the database with a Mac computer may not function properly (some functions work with Mac OS X operating system and Safari browser).

The mutation pattern option produces a pie chart showing the proportion of the different types of mutations (classified by their nature: base change, insertions, deletions....) in the selected set of tumors (calculated as follows: number of mutations of each type divided by the total number of mutations selected).
You can get a mutation pattern pie chart from the Basic search or Advanced search options. The Basic search allow you to select a tumor site (Topography) and/or a tumor type (Morphology) whereas with the Advanced search a precise set of data can be selected by defining the type of tumor, the population affected and the references in which the mutations are described.

Strand-bias
This option displays a table showing the description of the mutation event by mutation type for single base substitutions. This allows the calculation of the strand-bias (non-equal distribution of a mutation on each strand of DNA) for each mutation event.

The Mutation Effect option produces a pie chart showing the proportion of different types of mutations (classified by their effect: missense, nonsense,....) in a selected set of tumors (calculated as follows: number of mutations of each type divided by the total number of mutations selected).
You can get a Mutation effect pie chart from the Basic search or Advanced search options. The Basic search allow you to select a tumor site (Topography) and/or a tumor type (Morphology) whereas with the Advanced search a precise set of data can be selected by defining the type of tumor, the population affected and the references in which the mutations are described.

The codon distribution option produces a histogram showing the proportion of single base substitutions at each codon in a selected set of tumors (calculated as follows: number of single base substitutions at each codon divided by the total number of single base substitutions in the selected dataset).
The display option allows the setting of a minimum percentage value for a codon to be labeled on the graph. A 3D view of most frequent mutations within the DNA-binding domain is also displayed. This view requires that you download and install CHIME plug-in (unzip with the free software ZipGenius).
You can get a Codon distribution from the Basic search or Advanced search options. The Basic search allow you to select a tumor site (Topography) and/or a tumor type (Morphology) whereas with the Advanced search a precise set of data can be selected by defining the type of tumor, the population affected and the references in which the mutations are described.
Only single base substitutions within the coding sequence are included in this graph (selection of other mutations will return no results).

The Function pattern option produces a scatter plot showing the proportion of single base substitutions in a selected set of tumors, labeled with functional consequences and ordered according to their mutation rate. Different functional labelling can be selected: (1) transactivation activity on 8 different promoters in yeast assays (as reported by Kato et al.), (2) functional predictions by tow different methods (SIFT and AGVGD) that are based on residue conservation from fish to placental mammals, (3) functional predictions of missense mutations in the DNA-binding domain derived from structural features of mutants that are based on a four-body statistical potential scores calculated from the Delaunay tessellation of the wild-type p53 structure. The mutation rate for an amino-acid substitution is computed by taking the sum of the mutation probabilities of all individual nucleotide substitutions that lead to this amino-acid substitution. See mutation annotations for details on these functional classes and on the computation of the mutation rates.
The mutation frequency was calculated as follows: number of each specific mutant divided by the total number of mutants with available functional classification in the selected dataset.
You can get a Function pattern from the Basic search or Advanced search options. The Basic search allow you to select a tumor site (Topography) and/or a tumor type (Morphology) whereas with the Advanced search a precise set of data can be selected by defining the type of tumor, the population affected and the references in which the mutations are described.
Only missense, nonsense and silent mutations with mutation rates are included in these graphs (selection of other mutations will return no results).

The Tumor Spectrum option produces a histogram showing (1) for somatic mutations, the proportion of the selected mutations by tumor sites (percent of the total number of mutations present in the database for each tumor site), (2) for germline mutations, the type of tumors that have been reported in individuals who are confirmed constitutive carriers of the selected mutation(s) (percent of the total number of tumors reported in these individuals).

This option allows you to select a precise set of data by including and/or excluding several criteria. There are two search windows with the same criteria, the first one displays the 'criteria to include' and the second one the 'criteria to exclude'. There are four categories of criteria: criteria to describe the mutation, criteria to describe the tumor sample, criteria to define the patients and criteria to select references in which the mutations have been reported. The criteria for 'Mutation', 'Sample' and 'Individual' are displayed in the 'Advanced Search' window, whereas the 'Reference' criteria are displayed in a wizard, opened by clicking on the 'Add' reference button. Once you have selected the criteria to include and/or exclude, you can get the same type of graphs as for the basic search option.

Reference search
Click on 'Add' in the Reference criteria to open the reference search window. Enter a selection keyword (authors, journal name, title, publication year, ...). All references that include this keyword in the selected field will be displayed one by one in the wizard. Use the 'next' and 'previous' buttons to navigate between references. By default, all references are included. To exclude a reference, uncheck the check box. If the wizard was opened from the 'criteria to include' window, only the mutations described in the selected references will be included in the advanced search. If the wizard was opened from the 'criteria to exclude' window, the mutations described in the selected references will be excluded from the advanced search.

This option allows you to retrieve the TP53 gene status of cell-lines that have been screened for TP53 and reported in the literature or in the Sanger database. You may select any of the available criteria. A brief description of the retrieved cell-lines is displayed in the result window and detailed information can be downloaded.

Search tips: As different names or description may have been used (A431 vs A-431, or MDA-MB-231 vs MDA-MB231) for the same cell-line, it is not recommended to type the entire name of a cell-line, especially if it contains symbols such as - or /, . Use a partial name instead (431 or MDA), as the cell-line name criteria automatically includes wild cards on both sides of the input.

This option allows the search of functional activities of p53 mutant proteins. You can select a mutation or a group of mutations and get a table that shows the number of experimental data available for each category of functional properties. From this table, select a category to get a table displaying experimental data obtained in yeast or human cells. With the download option, you will get all functional properties of the selected mutations. Mutations that may not be easily selected with the available criteria (Complex mutations) can be selected with a scrollbar menu.

This option allow to retrieve the proportion of mutated samples by cancer type (topography or morphology) and by country of origin of the patients. You can select a type of cancer, a population and the method of mutation screening. You have to select the type of graph you want to display. The prevalences shown on the result graphs are calculated from the prevalence dataset (number of mutated samples divided by number of sample analyzed). Note that silent mutations are counted as mutation event in this dataset. Note also that mutation prevalence is highly affected by the method of detection, so please download and check the detailed information before interpreting graph results.

This option allows to select for a specific mutation and retrieve standard nomenclatures and database information for this mutation. Different input format can be used to select for a mutation. The information retrieved includes the precise description of the mutation at the DNA and protein level, observed (in experimental cell assays) or predicted (by amino-acid conservation rules, structural analysis, or splicing prediction) functional impacts of the mutation, number of occurence as somatic or germline mutation, and mouse-model with engineered p53 mutant.

To input the correct numbers for mutation position, please check the reference sequences:

This option provides tools to visualize primary and tertiary sequences of p53 wild-type or mutant proteins. For the 3D viewer and to display solved mutant structures, you need to download and install CHIME plug-in (unzip with the free software ZipGenius).
The Computational geometry analysis tools draw a histogram representing the Residual Score Profile of a selected mutant. Principle and methods of the Computational geometry analysis is available here.

The full datasets can be downloaded from the "Database downloads" pages and selected subsets can be downloaded from the graph pages with the "Database search" options. The 'Save as table' option allows to download the table containing the data used to generate the graph and the 'Download data' option allows to download the complete set of data linked to the selected data. Data are downloaded as tab-delimited text file compressed in a zip format. To open the zip file, use the free software ZipGenius. Graphs image can be saved using the mouse right-click button over the graph image. Description of the column contents are given in this help page.

The dataset of Somatic Mutation contains exclusively TP53 somatic mutations that have been identified in human tissues by sequencing and published in the peer-reviewed literature. This includes mutations found in normal, pre-neoplastic and neoplastic tissues, including metastases, as well as in cell lines derived from such tissues. The database does not include (1) individual data on human tissues that are reported as negative with respect to p53 mutation, (2) mutations not precisely identified by sequencing (e.g. mutations identified only by SSCP, DGGE or restriction digestion), (3) experimentally-induced mutations in tumor cells or cell lines in vitro, (4) p53 mutations in animal tumors.

TP53SomaticR14 is a compressed file in a zip format. It contains a Disclaimer and two tab-delimited text files: TP53SomaticR14 that includes mutation data and TP53SomaticRefR14 that includes the references in which are described the mutations.
To open this file, use the free software ZipGenius.

Each row in the table represents a single mutation with an arbitrarily assigned unique identification number. A unique identification number is also attributed to the tumor sample and to the patient.

This file lists the publications in which are described the mutations and gives the method used to detect the mutations. Each row (record) represents a citation with an arbitrarily assigned unique identification number (Ref_ID). See standardized annotations for the description of the column content.

The Prevalence Table provides information on the prevalence of TP53 mutation by tumor type.  It contains all publications having identified the mutations by sequencing. This includes publications contained in TP53SomaticR14Ref, but also publications that do not give a detailed description of the mutations (many recent studies do not provide detailed information on each mutation detected but rather report their results in the form of summary tables), and publications with negative results (no mutation found).
For each study, the total number of tumors or tissue samples analyzed, and the number of these samples which were found to contain a mutation are recorded. The reference paper, method of mutation detection and country of origin of the patients are also indicated. When the same research team published several papers that describe the same set of samples, data from the most recent paper are used.

The file TP53PrevalenceR14.zip is compressed in a zip format. To open this file, use the free software ZipGenius. It contains a Disclaimer, and a tab-delimited text file, TP53PrevalenceR14.txt.

A "Prognosis" information unit has been added to the Somatic Mutation Database since the R7 release. This feature consists in a table that includes information on all studies that have analyzed the relationship between p53 mutations and cancer prognosis. For each study, the patient cohort, study settings and a summary of the results are described. When the same research team published several papers with increasing number of patients, the most recent paper with the largest dataset is used.

Many of these studies do not provide detailed information on each mutation detected but rather report their results in the form of summary tables. Such publications have been included in the prognosis table, without mutation data attached (these references will not appear in the SomaticR14 file or through the database search system). For some of them, the mutations have been published in a previous paper and can be retrieved with the Cross_Ref_ID (see below).

The file TP53PrognosisR14.zip is compressed in a zip format. To open this file, use the free software ZipGenius. It contains a Disclaimer, and data either in a tab-delimited text file, TP53PrognosisR14.txt, or in a graphical output format in Microsoft Access, TP53PrognosisR14.mdb. The text file is compatible with softwares that handle tab-delimited formats, whereas the mdb file can only be opened with Microsoft Access 2000 software. Data in Microsoft Access format are easy to view and search with the filter function of Access forms (click here for help with the form filter function).

The table contains the following information:

TP53GermlineR14 is a compressed file in a zip format. It contains a disclaimer and two tab-delimited text files: TP53GermlineDataR14 including the mutation data and TP53GermlineRefR14 including the references.
To open this file, use the free softwareZipGenius.

Each row (record) represents a tumor found in an individual having a TP53 germline mutation. The columns contain the following information and abbreviations:

note: na= non applicable

Each row represents a reference identified by a unique identification number (Ref_ID). See standard annotations for a description of the column content.

The files TP53MUTfunction1R14 and TP53MUTfunction2R14 are compressed in a zip format. To open these files, use the free software ZipGenius. They contain a Disclaimer and a tab-delimited text file, TP3MUTfunction1/2R14.txt.

Data were extracted from publications that report functional assessment of p53 mutant proteins in human or yeast cells, either by transfection and overexpression of mutant proteins, or by assessment of endogenous mutants.

Comparison between mutants requires caution since functional assays differ from one study to the other, in particular with respect to the expression vector (which influences the level of expression of the mutant protein), the p53-responsive elements (generic consensus sequence versus gene-specific response elements from WAF1, BAX or PIG3), and the recipient cells that have been used. Indeed, p53 mutant functional properties appear to be dependent upon the type of cells and response elements.

The functional properties of mutant proteins that are included in this dataset are:
- transcriptional activities on various well-described p53-RE (see list here),
- dominant negative effects on the activities of wild-type p53,
- capacity to induce apoptosis, cell-cycle arrest or checkpoints in human cells,
- capacity to transactivate promoters that are not induced by wild-type p53,
- ability to promote cell growth and confer tumorigenicity,
- sensitivity to temperature changes regarding their ability to transactivate specific promoters.

The functional results have been organised in 5 columns for (1) conserved wild-type properties, (2) complete or partial loss of wild-type properties, (3) dominant-negative effects, (4) gain of function and (5) temperature sensitivity. The cell system is indicated in two columns and a detailed reference to the published report is given.

The columns contain the following information and abbreviations:

Note: NA= Non Applicable

The functional data that are included in this dataset were provided by Chikashi Ishioka and have been published in Kato et al., Kakudo et al., and Kawaguchi et al..

The columns contain the following information and abbreviations (for the detailed experimental procedures, please see the above references):

Note: NA= Non Applicable

The file TP53CellLinesR14 is compressed in a zip format. To open this file, use the free software ZipGenius. It contains a Disclaimer and a tab-delimited text file,TP53CellLineR14.txt, which is compatible with softwares that handle tab-delimited formats.

The current list of cell-lines include cell-lines that have been screened for TP53 mutation and are either published in the scientific literature or listed in the Sanger cell-line database. Cell-lines with a wild-type or mutated TP53 gene are listed.

The file contains contains the following information:
note: na= non applicable

The list of polymorphisms shown in the table includes gene variations that have been reported in publications or SNP databases and validated at least by frequency in unaffected human populations.

The table contains the following information:
note: na= non applicable

The table contains mouse models with engineered p53 that are compiled in the caMOD database or reported in the scientific literature. Data curated at caMOD were courteously provided by the caMOD team. A direct link to the caMOD web site is available for each model for a detailed description of model genetics and phenotypes. Data reported in the literature but not compiled in caMOD are curated at IARC and a link to PubMed abstract is provided. For a detailed description of model genetics and phenotypes, please refer to caMOD and/or original publication

The table contains contains the following information:
note: na= non applicable

Tumor samples are classified according to standards of the International Classification of Diseases for Oncology (ICD-O 3rd Edition, World Health Organization, Geneva, 2000) and SNOMED.
For information on tumor classification, grading and staging, check out ICD-O training at SEER, Collaborative staging initiative, Cancer Information at NCI, and Oncologychannel.com.

The same references (same Ref_ID) are used for the somatic, prevalence and prognosis data sets. Independent references are used for the Function and Germline data sets.