Nature Computational Science publishes primary research in the format of an Article. This format is not a review of technology, but instead an in-depth report on a relevant research study of high quality and general interest to the computational science community. This includes research on innovative algorithms and computational techniques that help advance science in one or more areas, as well as research on a scientific discipline that uses existing computational capabilities in novel ways to find new insights and solve challenging relevant problems. Regardless of the focus, to guarantee immediate practical relevance, Articles must show an application of the proposed computational technique to an important question or problem in a scientific discipline and demonstrate its advantage over existing approaches.
Articles begin with an unreferenced abstract (typically 150 words) and are divided into separate sections for Introduction (without heading), Results, Discussion and Methods. The Introduction section should describe the main problem being addressed, and is brief and focused; the Results section usually contains a general description of the computational technique or mathematical model being proposed or used, followed by its validation; the Discussion section summarizes the main findings of the research and clearly describes any existing limitations; the Methods section provides all technical details necessary for the independent reproduction of the results, without referring to a chain of bibliographical references. The main text (excluding abstract, Methods, references and figure legends) is 2,500–3,500 words. Articles have no more than 6 display items (figures and tables). The Results and Methods sections should be divided by topical subheadings. As a guideline, Articles allow up to 50 references.
Articles include received/accepted dates and may be accompanied by Supplementary Information. Articles are peer reviewed.
A Resource is typically used to report on and present a large dataset, a new tool, or a new framework of broad interest for the computational science community. The main text (excluding abstract, Methods, references and figure legends) is approximately 3,500 words. The abstract is typically 100–150 words and is unreferenced. Resources have no more than 6 display items (figures and/or tables). An introduction (without heading) is followed by sections headed Results, Discussion and Methods. The Results and Methods should be divided by topical subheadings; the Discussion does not contain subheadings. As a guideline, Resources allow up to 50 references.
Resources include received/accepted dates. They may be accompanied by Supplementary Information.
A Brief Communication is a more concise format used typically to report a significant improvement or adaptation to an existing research; to outline a novel, ongoing work (with some preliminary results) that holds out the strong promise of eventual high impact; or to anticipate new challenges in the field of computational science. Brief Communications begin with a brief unreferenced abstract (100 words). The title is limited to 10 words (or 90 characters) . The main text is typically 1,000–1,500 words, including the abstract, and may contain subheadings at the discretion of the editor. If the Brief Communication presents preliminary results, it must have a Methods section to provide all the technical details necessary for the independent reproduction of the methodology. Brief Communications normally have no more than 2 display items, although this may be flexible at the discretion of the editor. As a guideline, Brief Communications allow up to 20 references.
Brief Communications include received/accepted dates. They may be accompanied by Supplementary Information. Brief Communications are peer reviewed.
A Review is an authoritative, balanced survey of recent developments in a research field. Although Reviews should be recognized as scholarly by specialists in the field, they should be written with a view to informing non-specialist readers. Thus, Reviews should be presented using simple prose, avoiding excessive jargon and technical detail. Reviews are approximately 3,000–4,000 words long and typically include 4–6 display items (figures, tables or boxes). As a guideline, Reviews allow up to 150 references; citations should be selective. Footnotes are not used. The scope of a Review should be broad enough that it is not dominated by the work of a single research institution, and particularly not by the authors’ own work.
Reviews include received/accepted dates, and are usually commissioned by the editors. Reviews are always peer reviewed to ensure factual accuracy, appropriate citations and scholarly balance.
A Perspective is intended to provide a forum for authors to discuss techniques and ideas from a personal viewpoint. They are more forward looking and/or speculative than Reviews and may take a narrower field of view. They may be opinionated but should remain balanced and are intended to stimulate discussion and new experimental approaches. Perspectives follow the same formatting guidelines as Reviews. As with Reviews, many Perspectives are invited by the editors, so it is advisable to send a pre-submission enquiry including a synopsis before preparing a manuscript for formal submission.
Perspectives include received/accepted dates. Perspectives are always peer reviewed and edited by the editors in consultation with the author.
News & Views
News & Views articles inform readers about the latest advances in computational science research, as reported in recently published papers (in Nature Computational Science or elsewhere) or at scientific meetings. Unsolicited contributions will not normally be considered, although prospective authors are welcome to make proposals. News & Views articles are not peer reviewed, but undergo editing in consultation with the author.
The Correspondence section provides a forum for comment on issues relevant to the journal’s community. This format may not be used for presentation of research data or analysis. A Correspondence should not exceed more than two printed pages and can range from 300-800 words; it is limited to one display item and up to 10 references. Article titles are omitted from the reference list. Correspondence may be peer-reviewed at the editors’ discretion. Note that Correspondence pieces are not technical comments on peer-reviewed research papers which would be considered Matters Arising.
Matters Arising are exceptionally interesting and timely scientific comments and clarifications on original research papers published in Nature Computational Science. These comments should ideally be based on knowledge contemporaneous with the original paper, rather than on subsequent scientific developments.
For detailed information on how to submit a Matters Arising, please follow the instructions here.
Comment articles can focus on policy, science and society, or purely scientific issues related to computational science. Single-author articles are preferred as this is an ‘opinion’ section of the journal. Comments are usually commissioned by the editors, but proposals are welcome. They should be of immediate interest to a broad readership and should be written in an accessible, non-technical style. Figures and diagrams are encouraged, but are not a requirement. Comments are typically no longer than 1,500 words and, as a guideline, allow up to 15 references.
Comments may be peer reviewed at the editors’ discretion.