I- Scientific Background

General description / relevance and originality of the project Prognosis of a leukaemia depends on the genes involved: 5 years survival: 6% in the inv(3)(q21q26) RPN1/MECOM leukemia, 100% in the  dic(9;12)(p13;p13) PAX5/ETV6 leukemia. And treatments, indeed, depend on the severity of the disease. However, 2,000 to 9,000 genes are possibly implicated in cancer, and 1,200 types of solid tumours exist. Some cancers are frequent while many others are very rare (only 1 published case), this is particularly true for leukemias subtypes … but there are more than 900 leukemias!No one, anymore, has the whole required knowledge, necessary to guide the treatment procedure in case of a rare disease. Huge database are therefore needed, to collect and summarize data on these rare diseases, and produce meta-analyses. L’ Atlas contributes to 'meta-medicine', this mediation between the knowledge and the knowledge users in medicine.

II- International positioning of our team / Current organization

The Atlas is one of the pioneers of the Internet, it started in 1997. It meant writing short cards on genes and chromosomes implicated in cancer. The Atlas is now 45,500 pages big, written by more than 3,500 authors, from about 50 countries (France, USA, Italy, United Kingdom, Germany, Japan, Spain, Canada, China, The Netherlands ...).

The Atlas is in free access, which is particularly useful for third world countries or for students.

- Topic research The Atlas participates in research on cancer epidemiology. The Atlas is a tool in genomic studies of the latest generation (see Science 28 Jan 2011, vol 331 p 435-439, suppl data, where the authors indicate that they have used the Atlas and Cosmic as reference databases to choose and test a series of genes in medulloblastoma in childhood). Nonetheless, the Atlas' aim is not to overlap with the databases daily used in molecular biology (UCSC, Ensembl, Entrez, Cosmic, Swiss-Prot). However, as these data are in the public domain, we could easily produce them, according to the needs of the community.

- Topic teaching and continuing medical education The Atlas is at the crossroads of research, university and post-university teaching (virtual medical university) and telemedicine. 

- Database on genes: Where else can be found so many detailed articles on genes? see: PTN or BCL6 http://atlasgeneticsoncology.org//Genes/PTNID41904ch7q33.html and http://atlasgeneticsoncology.org/Genes/BCL6ID20.html.

- Electronic journal version of the Atlas An Open access electronic journal /pdf version of the Atlas is being developed by Institute for Scientific and Technical Information (INIST) of the French National Centre for Scientific Research (CNRS). Available are the archives of a quarterly journal since 1997, which became a bimonthly journal in 2008 and a monthly journal in 2009, comprising 2,500 articles in more than 120 volumes, which constitutes a 10,000 pages collection, available at: http://irevues.inist.fr/atlasgeneticsoncology, allowing the Atlas, soon, to be referenced by the main bio-medical databases, including PubMed.

- The Atlas combines various types of knowledge all on one site: Genes and their function, cell biology, diseases, cytogenetics, but also clinical genetics, including hereditary diseases which are cancer-prone conditions. This tends to unify cancer genetics, while data are elsewhere dispersed between several sites[1]. The iconography in the Atlas is diverse (medical imaging, pathology, chromosomes, 3-D structure of proteins, genetic maps...), which is not found in other sites (apart from genetic maps that can be found in GeneBank, Ensembl …). The Atlas is the only site devoted to genetics where the prognosis is quoted:  It has always seemed surprising to us that such a crucial data is just ignored in other sites. There are more than 17,000 internal hyperlinks in the Atlas.

- Diagnosis assistance and information-based therapeutic decision The Atlas contributes to the cytogenetic diagnosis and may guide treatment decision making, particularly regarding rare diseases (because they are numerous, rare diseases are frequently encountered).

- Developments in cell biology and physio-pathology: data in the Atlas are a definite resource in cell biology and physio-pathology, that we are just beginning to harness (e.g.: Apoptosis: http://atlasgeneticsoncology.org/Categories/Apoptosis.html ).

- Towards a personalized medicine of  cancer: From our section "Genes", can be extracted 613 genes implicated in colorectal cancer, 750 in breast cancer, and 494 genes in prostate cancer (see paragraph "Other genes implicated " at: http://atlasgeneticsoncology.org/Tumors/breastID5018#EXTRACTED). With the fast development of technics in genetics, it now emerges that many subtypes of solid tumors may exist, following the leukemia model (how many hundreds of breast cancer subtypes, defined by distinct genetic profiles, to be uncovered?). Recently, new data on lung adenocarcinoma made possible to consider personalized medicine (see http://atlasgeneticsoncology.org/Tumors/TranslocLungAdenocarcID6751.html ). This, together with cell biology developments in the Atlas demonstrates that the encyclopedic content is potentially a basis for developing personalized medicine for cancer. It remains to present this knowledge in such an affordable manner that the Atlas truly becomes a tool for the staff meetings. 

Renown Prestigious journals, such as Science and Nature Reviews Cancer, have written about the Atlas The Atlas is regularly cited as a journal in leading scientific publications (Annual Review of Biochemistry (Impact Factor 29,88), Science (29,75), Nature Reviews Cancer (29,54), Cancer Cell (25,29), Nature Cell Biology (19,53), Journal of the National Cancer Institute (14,07), American Journal of Human Genetics (12,30), Molecular Systems Biology (12,13), Genes and Development (12,08), Genome Research (11,34), Trends in Molecular Medicine (11,05), Blood (10,56), PNAS (9,43)...). The Atlas is also cited as a reference database in Science. Hyperlinks towards the Atlas have been developed by the National Cancer Institute USA, Swiss-Prot, GeneCards, the Sanger Institute, Mitelman database au NCI, etc... Renown of the Atlas may be checked at: http://chromosomesincancer.org/en/jce/acknoledgements-to-the-atlas.html (Felix Mitelman wrote that the Atlas "has grown into a truly monumental encyclopaedic work of great importance to cancer research (...) an invaluable reference and resource for scientists and clinicians").

Who uses the Atlas?The Atlas is accessed by: 1- Research: cytogenetics, molecular biology, cell biology researchers; 2- Hospital: clinicians, haematologists, cytogeneticists, pathologists, from the teaching hospitals, indeed, but also from general hospitals where the Atlas is one of the rare free resources. Junior doctors in haematology or oncology, are also most receptive to the Atlas that they see as a training and educational tool; 3-Students.

The Atlas receives more than 1,100,000 visits every year (25% of the visitors are from the USA, 12% from France, but many other visitors are from the developing countries. 200 cytogenetic labs come every day, 1,000 visitors more than once a week (cytogenetic labs./university teachers-researchers), and 6,000 visitors more than once a month (university teachers - researchers, students, staff meetings on blood malignancies). 5,500 individual machines connect to us every day. A bug, on May the 23rd, 2008, was devastating: pages on chromosomes, with their data, had vanished; many mails were sent to us, proving that the Atlas was (is) indispensable; see: http://chromosomesincancer.org/en/jce/acknoledgements-to-the-atlas.html#Incident ). See also recent testimonies at :http://atlasgeneticsoncology.org/Supporting_Atlas_First_signatories.pdf

The Atlas is also operated by various biotech and pharma to implement their internal systems and databases.

Comparison with the Mitelman Databasethe Atlas is certainly not redundant with the "Mitelman". As a matter of fact, the Mitelman and the Atlas are complementary: the Mitelman is exhaustive but gives rough data, with no annotation, while the Atlas presents a meta-analysis of the data, giving an overview on a given disease, describing the main clinical characteristics, with, when possible, an iconography of chromosomes (e.g. see: t(1;11)(p32;q23) in each of the two sites).

Scientific societies have decided to grant the Atlas, they thus give a scientific credibility and a clear and real will to the Atlas sustainability and perennialty: Association des Cytogénéticiens de Langue Française et Groupe Francophone de Cytogénétique Hématologique, Belgian Society of Human Genetics et Belgian Cytogenetics Group in Haematology and Oncology, Dutch working group of Tumor Cytogenetics, Berufsverband Deutscher Humangenetiker e.V. et Gesellschaft für Humangenetik e.V., Societa Italiana di Genetica Umana, Grupo Cooperativo Español de Citogenética hematológica and Sociedad Española de Hematología y Hemoterapia, Australasian Society of Cytogenetics. 

Structure of the board of the Atlas

- Jean-Loup Huret, associate professor and consultant, CHU Poitiers, editor in chief and Philippe Dessen, research director, CNRS-Institut Gustave Roussy, database director;

- Because, historically, the project could not have been supported by institutional players, a non-profit association (french law 1901) was created (see http://chromosomesincancer.org/en/ ) (president Jean-Loup Huret, geneticist, vice president Hossein Mossafa, geneticist, treasurer Bernard Drochon, external auditor, secretary Martine Jammet, entrepreneur, all volunteers). This association have employed 5 to 6 MSc or PhD co-workers (the main area of expenditure: 30,000 Euro (salary + charges) for each collaborator, but they make an indispensable work -since the Atlas is much more an encyclopedia than a database, requesting a good amount of expertise). The association comprises about 60 members (50% french, 25% americans).

- A scientific committee of the association (http://chromosomesincancer.org/fr/jce/conseil-scientifique.html ), including delegates from various scientific societies; and there is also an editorial board for the Atlas (http://atlasgeneticsoncology.org/Backpage.html#EDITORIAL ).

III- Development steps

To reach its principal object -the patient service-, the Atlas must become "clinic orientated". To encourage the clinical physician to use the Atlas in his decision making, the Atlas must provide the required information, and such in a "clinician friendly" manner.

A part of our activity must be dedicated to bioinformatics developments, at a time when new fields of cytogenetic are developing with the massive use of FISH, CGH, exome and genome sequencing. Numerous and precise data on mutations, structures variations as well as on gene fusions are taking over the literature.

Adding information with a clinical orientation and into a customized therapeutic care. Developing the knowledge and the scientific innovation transfer into the care system, in particular as part of targeted therapies. Developing decision trees (using the High Authorities recommendations for each pathology); making a list of all the examinations to be performed according to each step of the

Editorial workflow in the Atlas

The Atlas editorial team stands in need of Section Editors (see http://atlasgeneticsoncology.org/BackpageAbout.html#EDITORIAL ) devoting 10% of their time for the Atlas.

The workflow detailed herein below applies when MSc students/workers are involved in the process. Indeed, a senior researcher will skim through the various steps described below. However a secretary's active watch on forthcoming papers from the various authors is needed (like an appointment calendar). Moreover, when the Atlas has evolved into a true database with an automated editorial process, many of these time consuming tasks will vanish.

For each of the themes included in the atlas (genes, leukemias, solid tumors, hereditary diseases involving increased risk of cancer, "deep insight" …), the editorial team is tasked with appraising the subjects (or "items") lending themselves to publication on the site.

As regards genes, for instance, using  a predefined list of the genes potentially implicated in cancer (a list determined by algorithms for different data bases by Philippe Dessen, the data base manager), the editorial team is initially tasked with verifying the implication (or non-implication) of these genes in carcinogenesis. With this in mind, they study the existing bibliography, primarily on the basis of the information contained in PubMed. They then endeavor to determine whether or not the published data suffice to justify a detailed written review.  

As regards the other themes (leukemias, solid tumors, hereditary diseases involving increased risk of cancer...), the bibliographic research is also carried out from PubMed, and also on specialized sites (Mitelman, Orphanet, OMIM, WHO, …). This research phase favors subjects for which enough recent publications are available to allow for high-quality reviews.

Finally, the editorial team decides whether or not a given subject will constitute an item justifying publication. After that, it is a matter of searching for and selecting the author in the best position to write this review.

Once a subject has been determined, the editorial team conducts a search of recent publications in view of choosing the author/the team in the best position to write something for the Atlas: their second appraisal. Concerning their choice, the following factors have got to be taken into account: number of publications by the author, the impact factor of his or her publications, and the author's status: clinician, hospital-based biologist, researcher. The author's disciplines are likewise taken into consideration: oncology, a specific organ, cytogenetic biology, molecular biology, cellular biology, physiopathology, etc. As regards teams, a multidisciplinary team will be preferred.

This is a highly sensitive step. According to the author chosen, the quality of the review is likely to vary, and its scientific orientation will be more or less fundamental or clinical.

Other selection criteria include: qualities of clarity in data presentation, synthesis and the use of schemas or medical imagery (in the broad sense of the word, including anatomopathology and chromosomes).
 

Along with the different items (genes, leukemias, solid tumors...), all important data on the authors are conserved in dedicated files.

And for each gene, the spreadsheets inventory information on its implication, as well as all other information deemed useful for future research (authors to be contacted, their contact information, e-mail exchanges, agreement to write an article, speed in production, and subjects to be reserved for a given author...). Totally indispensable, the spreadsheets are of concrete use during each phase of the work. They need to be ergonomic, and evolve according to changing needs. 

For each theme in the Atlas (Genes, Leukemias, Solid Tumors, Hereditary Diseases involving increased risk of cancer, "deep Insights", Teaching Chapters), there exists a dedicated spreadsheet.

Each spreadsheet provides a list of the subjects studied, the relevant dates, as well as the subjects to be studied, for which the right authors remain to be found.

The spreadsheets are updated daily, and new, potentially interesting subjects are added periodically. This supplementary material is enriched by readings of articles from the literature and by analysis of the articles received by the Atlas.   

Each subject is designated by a specific and numbered identifier facilitating the establishment of internal links as the files are being edited.

Possible duplicate files are regularly sought out and subjected to analysis, and the spreadsheets are consequently "cleaned".  

For an example, see below: Extracts from the "Genes" spreadsheet, which contains 9000 lines.

The authors are contacted through a common e-mail account using predetermined contact e-mail coordinates.

The e-mails are sent once a week, at the beginning of the week, as it has been observed that the highest response rate occurs at that time. On the same token, some times of the year are not conducive to soliciting authors who tend to be busy with applications for grants ... or on holiday. These periods are avoided as much as possible.

When an author fails to answer, he is re-contacted two or three times. If, after that, he still fails to answer, another author is sought out...

Author response statistics: 55% of the authors do not respond to the invitation; 25% of them refuse; 20% agree.

When authors refuse or do not respond, the process of searching for the right author is reinitiated. If no other author can or will write the requested review, the team will wait for new research to be published..

When, on the other hand, an author accepts, the team consults him on the time frame he thinks he shall need prior to sending his completed article. The publishing flow is thereby managed, both for the Internet site and the PDF journal. A model of writing for the review is given the author, who concomitantly receives relevant editorial information.

Quite frequently, articles fail to arrive within the allotted time. The status of ongoing articles has got to be regularly monitored, and reminder notes need to be sent. 

It also often happens that notwithstanding an initially positive response, authors do not respond to reminder notes, or else they withdraw from the project. In those cases, search for the right author resumes.

Assistance for authors: Quite frequently, authors ask questions (deadlines , explanations pertaining to various details...) that call for timely responses.

 Once the article has been received, the completed worksheet is registered on a common server.

The information is also recorded on common spreadsheets, in a common notebook dedicated to the editors' ongoing activities; it is also displayed on wall calendars providing overall perspective on publication flows.

With these calendars, the editorial team can rapidly apprise itself of a possible lack of articles for the PDF journals and try to optimally manage the flow of arriving articles. Unfortunately, given the fact that the authors are the ones who determine their sending dates, substantial variations in publication flow may come to exist, and not be easy to manage.

The articles received are reread by internal referees; 97% of them (in fact all the articles, except for clinical case reports) are reviews of the literature written at the request of the Atlas editorial board; they are known as "commissioned papers". It should be noted that the authors were preselected according to rigorously applied criteria preliminarily to their writing of an article. Following receipt, the process of rereading and assessment of articles is implemented by the Atlas editorial board. It should also be noted that as the process draws to a close, the board member having chosen the subject (and subsequently the author himself) possesses a precise overall vision as to what the article is supposed to produce (quantity of knowledge in molecular biology or clinical research on a given gene...).     

Articles that do not correspond to the publication criteria or that do not provide a sufficiently detailed report on the knowledge available on the subject are sent back for revision. The editorial team remains vigilant, and about 20% of the articles submitted are indeed sent back for revision. It is often necessary to send reminders to the authors, before receiving the revised version (loss of time).  

The clinical case studies in hematology (the remaining 3%) are subjected to an external assessment process involving 4 or 5 experts, while the final decision is made by the Atlas editor-in-chief. Practically all the clinical case studies are reviewed by the authors at least one time prior to publication; only 8% are immediately accepted, 68% are accepted following revision, and 24% of the articles having been submitted are finally refused.
 

Once the articles have been accepted, they are processed so that they can be placed on line by the data base manager.

The articles are to be formatted according to a strictly applied computational model in order to be properly handled by the scripts of the data base manager (see, as an example, the form below).

The text must be painstakingly reread so as to correct possible mistakes (typos, grammar, spelling...),   and each bibliographical reference cited in the next has got to be verified and corrected (if need be).

The article header (the different tags or tracking devices allowing for indexing of the form in different parts of the data base in accordance with different criteria) must be carefully filled out, using the metadata contained in the article. The metadata will permit the data base manager to integrate the form within the site so that it can be inventoried with regard to the dedicated lists.  

The articles in the Atlas must follow a predetermined model; each paragraph of the review must be included within the tags, and the text must be formatted according to a computer encoding system. Special characters (accent marks, Greek letters, symbols…ex: integrin ανβ3) must be coded in the HTML format so that their posting on the website be correct. 

In the text itself, all potential internal links to other items contained in the Atlas (genes, leukemias, solid tumors...) must be identified, and internal links created in such a way that readers are directed, if they wish, to other articles in the Atlas. New subjects for the Atlas are often found while the forms are being processed, and added to the shared spreadsheets.  

Most often, the bibliography contains 50 to 100 references, and it has got to be processed in a predetermined format.

The articles also contain images, which are processed by Photoshop software. More often than not, the size of the images requires rectification, and the background of the image has got to be rendered transparent. Moreover, the text may need rewriting when it is too blurry or has been degraded due to the change in size.  

Once processing of the worksheet has been completed, it is put through a simulator so as to verify the correctness of its formatting, and also so as to make sure that no component remains invisible, as happens when a closing tag such as > is forgotten. 

The file and the attached images are then sent for validation to Jean-Loup Huret, the editor-in-chief, and subsequently to the data base manager for uploading or posting on line.  

"Deep Insight" and "Educational items" constitute special cases. As these files are "free format" and do not correspond to a precise model, they are directly processed using specialized software (the Dreamweaver website-editing software) in the HTML format. For these files, it is necessary to carry out all coding and page layout in HTML.

Example below : Partial formatting of an article: In the example given below, any and all parts of the text in color are subjected to special processing.

BEGIN_HEADER

FILENAMESH3PXD2AID45995ch10q24.txt

CLASSE GENE

ID45995

LOCUSID

TRI_PAR_CHROMOSOME10

TRANSLOC t(10;10)(q24;q24) SH3PXD2A/OBFC1

TRANSLOC t(10;13)(q24;q14) SH3PXD2A/RB1

FUSION_GENESH3PXD2A/OBFC1

FUSION_GENESH3PXD2A/RB1

CATEGORYCytoskeleton

END_HEADER

etc..

etc...

FUNCTIONTKS5 was initially identified as a substrate for <CC: TXT:SRC ID: 448> (Lock et al., 1998), and was subsequently shown to play a critical role in invadosome formation in multiple cell types (Courtneidge, 2011; Murphy and Courtneidge, 2011; Paz et al., 2013). <BR>

FUNCTIONFull-length TKS5 functions as an adaptor for recruiting other proteins to the cell membrane for invadosome formation. The recruitment of TKS5 to the cell membrane depends on its PX domain and phosphorylation by Src (Abram et al., 2003). It has been proposed that phosphorylation of TKS5 releases its PX domain from intramolecular interaction and allows TKS5 to bind to cell membrane phosphatidylinositol lipids, such as phosphatidylinositol-3,4-bisphosphate (PI(3,4)P₂) (Abram et al., 2003; Oikawa et al., 2008). At the cell membrane, TKS5 is thought to interact with multiple components of invadosomes either directly or indirectly, and thereby mediates invadosome formation and maturation (Sharma et al., 2013). These interacting partners includes adaptor proteins and actin regulatory proteins, such as <CC: TXT:NCK1 ID:41505>, <CC: TXT:NCK2 ID:52582>, <CC: TXT:GRB2 ID: 386>, <CC: TXT: CTTN ID: 369> (Cortactin), <CC: TXT: WASL ID: 42803> (N-WASP), <CC: TXT: ACTR2ID: 49744<CC: TXT: ACTR3ID:46303> (Arp2/3) complex, and <CC: TXT: ARHGAP35 ID: 52237> (p190RhoGAP) (Crimaldi et al., 2009; Oikawa et al., 2008; Stylli et al., 2009).<BR>

FUNCTION TKS5 also interacts with <CC: TXT:NOXA1 ID:41563> and <CC: TXT: CYBA ID: 43882> (p22phox), which are components of the NADPH oxidase complex, and thereby promotes reactive oxygen species (ROS) production by NOX enzymes at invadosomes (Diaz et al., 2009; Gianni et al., 2010; 2009).ROS have been shown to facilitate invadosome formation by maintaining or amplifying the phosphorylation of TKS5. As such, TKS5 is thought to promote invadosome formation via ROS in a positive feedback loop.<BR>

FUNCTION Finally, TKS5 has also been shown to interact with members of the ADAM family metalloproteases, specifically <CC: TXT:ADAM12 ID:44084>, <CC: TXT:ADAM15 ID:46345>, <CC: TXT:ADAM19 ID:46837>(Abram et al., 2003). It is believed that Tks5 recruits theses proteases to the invadosome foci for processing growth factors and regulating cell motility. For example, ADAM12 has been shown to promote ectodomain shedding of <CC: TXT:HBEGF ID: 40369> (heparin-binding EGF-like growth factor) and enhance invadopodia formation in cancer cells (Diaz et al., 2013). <BR>

etc...

Article upload takes place virtually every week, and for each article, it has to be verified.   Also to be verified: Has the article been correctly indexed in  the dedicated lists? Has the article been correctly displayed? .

Before a thank-you e-mail is sent to the authors of a review, their bibliographies are searched so as to see if they could possibly be tasked with another review. If this is the case, they are asked whether or not they would be interested in pursuing their collaboration.   

The author is given the address at which he can find his review on the Atlas site.

Prior to each issue, the editor in charge of the journal is tasked with creating the table of contents and scrupulously selecting the articles to be published. This is because publishing workflow is highly fluctuating, and consequently demands "hands-on" management.

A tool for the future data base carrying out initial formatting of the articles and editing the table of contents provides assistance in the page layout of the PDF journal.

In each issue, page layout is a sensitive matter. It is performed article by article, and has got to be done in such a way that a given issue is harmonious and satisfactorily balanced; moreover, image management must ensure maximal quality.

Once the journal is formatted and paginated, it is validated by the editor-in-chief.

The issue is then sent to the INIST-CNRS for finalizing of the PDF. The INIST editors recover the meta-data for the articles and associate them with the issue. They are also tasked with assigning the DOI (digital object identifier) for each article that will serve as its single identification number.  

For back-up purposes, this final PDF is registered on the future data base.

Once all of the above steps have been carried out, the authors are contacted anew and informed of the publication of their articles in the journal and the link to the latter in PDF with which they are being provided.  

 The above percentages are subject to modification according to editorial needs and the fluctuating importance of certain tasks.

Again, we want to make it clear that the workflow detailed herein applies when MSc students/workers are involved in the process. Indeed, a senior researcher will skim through the various steps. However a secretary's active watch on forcoming papers is needed. Moreover, when the Atlas has evolved into a true database with an automated editorial process, many of these time consuming tasks will vanish.

The Atlas can run with:

- an Editor in Chief (20% full-time) + secretary (2-5%).

- Section Editors (5- 10% full-time) + secretary (1%).

- a Bio-computer specialist (10% full time).

- .. and as many good authors as possible !

Added value of the Atlas

Encyclopedia containing original monographs written by different authors, the Atlas combines different types of knowledge. No other web site presents so many monographs on genes, very rich iconography. It is a tool for researchers in genomics, for clinicians as a help in diagnosis and therapeutic decision, for a personalized cancer medicine. The Atlas is not only a web site and an encyclopedia, but also a scientific journal.

In conclusion, it is an original/unique database without equivalent.

Positive aspects of the Atlas

Encyclopedia composed of documents that are original monographs written by invited authors, based on their expertise in a given field (ex: Carlo M Croce on "Common fragile sites and genomic instability", Rolf Marschalek on the KMT2A (MLL) gene, George C Prendergast on the IDO1 and 2 genes, Cécile Badoual on "Head and Neck paragangliomas", Maurizio Genuardi on "MUTYH-associated polyposis", Felix Mitelman on "Cancer Cytogenetics"). Other databases present texts on genes (for example Aceview) but these are non-expertized texts, resulting from data mining. GeneCards is a database with links to other databases, with few data copied from these databases.

Starting first from cytogenetics, the Atlas combines different types of knowledge in a single web site: genes and their function, cell biology (ex: Apoptosis:http://atlasgeneticsoncology.org/Categories/Apoptosis.html), pathological data, diseases and their clinical implications, cytogenetics, but also medical genetics, with hereditary disorders associated with an increased risk of cancer. This gives a wider and more global view of cancer genetics, while these data are usually dispersed[1]. The Atlas is the only genetic site where the prognosis is included; it has always surprised us that data so important such as prognosis could be absent in other sites… even more now in the context of personalized cancer medicine.

No other site, journal or book has so many monographs on genes (more than 1,400 genes; ex:http://atlasgeneticsoncology.org//Recent.html), together with 28,000 non-annotated cards based on the GeneCard type (ex: http://atlasgeneticsoncology.org//Indexbychrom/idxa_11.html, see paragraph on "Other genes").

Very rich iconography on chromosomal aberrations (700 images, ex:http://atlasgeneticsoncology.org/Anomalies/t0609ID1014.html), on genes (3,000 images ex:http://atlasgeneticsoncology.org/Genes/EWSR1ID85.html), on clinics and pathology (400 images, ex:http://atlasgeneticsoncology.org/Tumors/OralMelanomaID6647.html).

Research axis: the Atlas is a tool for researchers in genomics (see Science, 2011). Documents on genes having more and more importance in the analysis of an abnormality, mutations and fusions in relation with NGS: the Atlas is a tool allowing an expert selection of genes involved in cancer. It has been referenced several times as a pertinent source in relation with an abnormality, more particularly as identification of genes associated with the abnormality.

Help in diagnosis and therapeutic decision: the Atlas guides the cytogeneticist by its iconography on chromosomal abnormalities, informs the clinician on rare pathologies, allowing him to adjust his treatment. The Atlas is particularly indispensable in rare diseases (because rare diseases are numerous, these are frequent). All the cytogeneticists from all around the world regularly use the Atlas, as it appeared with recent financial issues[2].

Towards a personalized cancer medicine: 322 genes involved in colon cancer, 457 in breast cancer and 589 in prostate cancer can be extracted from the "gene" section (see paragraph "Other genes implicated"http://atlasgeneticsoncology.org//Tumors/breastID5018.html). With the development of new techniques, it is now evident than solid tumors, as leukemia, have numerous sub-types (how many hundreds of breast cancer with different genetic profiles?). This, added to the development in cell biology, shows that the encyclopedic contents of the Atlas is potentially a basis for the development of personalized cancer medicine.

The Atlas is not only a web site and an encyclopedia, but also a scientific journal. The pdf format of the Atlas (http://irevues.inist.fr/atlasgeneticsoncology) constitutes the archives of a journal published 4 times a year since 1997, then 6 times in 2008, and 12 times in 2009. It includes 2,500 articles in more than 100 volumes, and 9,000 pages to allow us to be referenced by PubMed in the future.

Ergonomy: Access by pages (ex: chromosomeshttp://atlasgeneticsoncology.org/Indexbychrom/idxa_14.html) , by table of anomalies, genes and fusion genes by chromosomal band (ex:http://atlasgeneticsoncology.org/Bands/11q23.html#GENES), with the more recent data on chromosomal abnormalities and fusion genes compiled (TCGA, Cosmic, TicDB, etc.). Synthetic tables (ex: nosologyhttp://atlasgeneticsoncology.org/Tumors/Solid_Nosology.html, patterns, categories, etc.).

The initial structure (independent documents) was improved to allow numerous internal hypertext links: i) direct links and ii) call for links from a given document to a wide range of other documents (ex: tumors associated with a given gene). The Atlas contains more than 17,000 internal links. It is bound to get enriched from its own contents by meta-analysis scripts.

Gestion of external links towards the majority of interesting public databases is automatic.

The Atlas is referenced by other web sites (NextProt, GeneCards, Mitelman database, COSMIC, etc.).

The Atlas benefits from the material support of INIST (CNRS-INSERM), allowing an excellent internet access (more than 1.5 million visits in a year).

In conclusion: original database with no equivalent or concurrence, therefore necessary, besides several other databases, for an in-depth analysis (ex: Atlas + Mitelman + Gene at NCBI + Cosmic).

… However …

Lack of exhaustivity and lack of update of the existing documents (due to lack of staff and the amount of work to be done - the Atlas has already more than 50,000 pages).

Need for the database restructuration, with an editorial system and a more performing ergonomics of consultation.

A need, very soon, for new developments concerning new techniques in genetics.

Note: Some addresses

- New developments:http://atlasgeneticsoncology.org/News.html

- Atlas Journal Pdf/Scientific side of the Atlas:http://documents.irevues.inist.fr/handle/2042/15655

e.g. http://documents.irevues.inist.fr/bitstream/handle/2042/56473/vol_19_4_2015.pdf

- Cell Biology: e.g.http://atlasgeneticsoncology.org/Categories/DNA_repair.html

- Cell lines:http://atlasgeneticsoncology.org/cell_lines.html

- Nosology: http://atlasgeneticsoncology.org/Tumors/Solid_Nosology_9.html

- Physical maps of genes, markers, and diseases at the level of the chromosome band: e.g. http://atlasgeneticsoncology.org/Bands/21q22.html

To all Cytogeneticists, Geneticists, Clinicians, and Scientists for whom the Atlas of Genetics and Cytogenetics in Oncology and Hematology still represents a cultural and practical tool ...

see: http://atlasgeneticsoncology.org//Supporting_Atlas_First_signatories.pdf