Peer review is the heart of any solid biomedical journal, and Blood is no exception. Each reviewer is asked to evaluate the work in detail and then rank the work on a number of points, the 3 most important of which are technical soundness, novelty, and the degree to which the findings advance the broad field of hematology. The development of hot new systems biology tools has provided unparalleled opportunities to hematology investigators worldwide. However, because the nature of those technologies is often hypothesis-generating rather than hypothesis-testing, the technologies have challenged some of our notions of this journal's role in advancing hematology research.
For example, an ideal gene expression microarray, proteomic, comparative genomic hybridization, or other global high-output data platform would be used to craft a paper, one that all Blood editors and most of our reviewers would agree should be published in the journal, that used well-designed, technically high-quality, and perfectly controlled experiments, and that resulted in a field-advancing outcome. The latter can be of 2 types. First, gene expression or some other high-density data platform signature can be shown to be of diagnostic or prognostic value. Second, individual molecules, genes, or genomic loci (or the signaling molecules that control their expression) are identified through the analysis, and then shown to have an important biologic or pathophysiologic role. Either of these outcomes often requires time, access to evolving clinical or biologic data, and extensive validation. Naturally, there are delays from the development of the data set to the confirmation of a key pathway or the demonstration that a profile has clear clinical value.
We have previously consistently rejected work that is less well validated. Even technically sound, well-designed studies have been turned down if they provide only potentially valuable findings. Generally the discussion sections of such manuscripts list ontologic categories to which the authors impute potential significance and narratives of why they believe the categories to be likely of great importance. The Editor-in-Chief and Associate Editors have agreed in the past that these kinds of strictly descriptive studies do not meet standards for publication in the journal because they fall short on the “field advance” score. That is, the manuscript describing the work itself has not yet reported a validated finding that advances basic or clinical hematology.
We also recognize a downside to this hard-nosed policy. That is, some data from studies that are truly novel and even inspired on the score of experimental design and quality, if released publicly, could truly advance the field by being available for use by others in the field. Each ontologic category and each candidate pathway identified in a well-done systems biology experiment could be of unique value to unique groups of investigators around the world, some of whom might not otherwise have ready access to the technology. For that reason, some of us believe that the best of these ought to be, somehow, the provenance of Blood.
In the past, we had expected that public repositories such as the Gene Expression Omnibus would serve to provide such data in a usable way, and in large part they do. Yet users of data in such sites often feel “half empty” because much of the background information that underlies the rationale for the study can be rather opaque. For well-designed hematology-focused studies, it is our view that such information (more than simply annotations) ought to be fully developed and reviewed. For this reason we propose to carefully review and publish novel systems biology work in 2 ways. For those that have fully validated their “hits” in functional ways (at the bench or at the bedside), nothing will change; these papers will continue to be published in Blood as Regular Articles. Those that are well designed and provide novel and extremely useful data but are strictly descriptive will be considered instead for publication in the new online-only section of Blood termed “e-Blood” as fully PubMed-citable articles.
Blood will not host the raw data, but e-Blood articles will provide links to the public repository sites and, more importantly, will provide concise narratives on the background and rationale supporting the design of the experiments, give details of the experimental design, and provide short discussions on the potential use of the data. We feel that this new category of scientific work, to be published only in the electronic version of the journal, will support the more rapid use of well-developed data by many groups and will facilitate more rapid advances in the field of hematology. More details regarding submission to this new feature of Blood can be found in the Instructions to Authors on the journal's website or by contacting the Blood editorial office.