Skip to main content

Archived Comments for: Annotation and query of tissue microarray data using the NCI Thesaurus

Back to article

  1. Mapping with The Developmental Lineage Classification and Taxonomy of Neoplasms

    Jules Berman, Consulting Editor, Jones & Bartlett Publishers

    6 November 2007

    The article by Shah and coworkers uses the NCI Thesaurus to map cancer diagnosis terms (for TMA tissue samples). I am writing this comment to suggest that the authors should have tried The Developmental Lineage Classification and Taxonomy of Neoplasms (hereinafter called the Developmental Classification), which has many more names of neoplasms (perhaps 10-fold more) than the NCI Thesaurus, and which is distributed under a GNU Free Documentation License.

    The Developmental Classification has a single biological organizing principle that classifies neoplasms by their lineage. There are six major classes of neoplasms:

    Endoderm/Ectoderm

    Mesoderm

    Neuroectoderm

    Neural Crest

    Germ cell

    Trophectoderm

    Each neoplasm occurs in only one subclass, and its lineage can be easily traced upward through subclasses until it reaches one of the six major classes. The rationale of the classification is that tumors inherit key cellular pathways through their developmental lineages. This assertion is supported by decades of morphologic evaluations of tumors. More recently, molecular biological observations have shown that genetic markers and pathways are carried through cell lineage. Tumors grouped by cell lineage may share responses to new chemotherapeutic and chemopreventive agents targeted to specific pathways. If this is true, we can start to develop agents (and combinations of agents) that are effective against groups of neoplasms that share a common developmental lineage.

    This theme has been developed in several of my early papers, one of which has had over 12,000 downloads from the BMC site and is the second-most cited paper of all-time in BMC Cancer.

    Berman JJ.

    Tumor classification: molecular analysis meets Aristotle.

    BMC Cancer. 2004 Mar 17;4:10.

    The classification was designed for fast parsing and has been prepared in three versions: an RDF ontology, a flat file (with the complete lineage of each term) and as a plain XML file.

    The gzipped version of the RDF file (under 1 Megabyte).

    http://www.julesberman.info/neorxml.gz

    The flat file version, listing each term followed by its lineage (gzipped file).

    http://www.julesberman.info/neoself.gz

    The plain XML version, with no RDF semantics (gzipped file). http://www.julesberman.info/neoclxml.gz

    The Developmental Classification contains several parts:

    1. The neoplasm classification proper (5,841 classified types of neoplasms and 130,503 terms representing the 5,841 types of neoplasms)

    2. A listing of cancer descriptive terms that are not names of neoplasms

    3. A listing of hyperplasias or hamartomas, some of which will be entered into the proper classification and others of which will remain in class Hyperplasia

    4. A listing of precancer terms

    5. A listing of inherited syndromes associated with increased risk for cancer.

    Classified plus unclassified terms exceeds 146,000 unique entries.

    The classification is closely curated, and frequent updates are announced on my blog site, Specified Life, devoted to medical annotation.

    http://julesberman.blogspot.com/

    It is not too late for the authors to try a neoplasm classification and taxonomy that was specifically designed for projects such as theirs. The Developmental Classification may not have the cachet of an NCI-produced project, but scientific progress would halt if every effort exclusively relied on materials provided by major organizations and agencies.

    - Jules Berman, Ph.D., M.D.

    Competing interests

    I have no competing interests.

Advertisement