Guidelines for the Reproducibility of Biophysics Research
Updated July 2015
The Biophysical Society, publisher of Biophysical Journal (BJ), agrees fully with the intent of the National Institutes of Health’s Principles and Guidelines for Reporting Preclinical Research to encourage reproducible, robust, and transparent research. The two basic principles underlying these Guidelines are the following:
First, research results should be reported with sufficient clarity and detail to ensure that the study can be replicated in any laboratory.
Second, data and material produced during the research leading to a published study should be readily disseminated and openly accessible, whenever feasible (e.g., by depositing the information in a public, community-endorsed database [see below], by including it in the online Supporting Material accompanying the article, or by posting it on the author[s]’ website).
By submitting a manuscript to BJ, the authors agree to adhere to these fundamental principles.
Adherence to these principles requires attention to the points listed below, where the specific guidelines will continue to evolve as new community standards emerge.
1. RIGOROUS STATISTICAL ANALYSIS
Any statistical analysis must be checked for accuracy by the authors; if statistical software is used, the source (including version number) of the tools must be listed in Materials and Methods. Authors are strongly encouraged to seek the assistance of a statistician at their institution or elsewhere; if they seek such advice, the resource should be identified in the letter of submission.
Statistics should be fully reported in the manuscript/article, including the statistical test used, exact value of N and the definitions of center and dispersion and the precision measures (e.g., mean, median, SD, SEM, confidence intervals). Manuscripts that report results based on the analysis of large data sets, such as Genomic, GWAS and NextGen Sequencing-based studies are required to specify in detail how the statistical analyses were done.
2. TRANSPARENCY AND REPRODUCIBILITY
BJ has no page limitations for research articles and allows for the submission of online Supporting Material. To ensure that published studies can be replicated, authors should adhere to the following standards:
- Describe Materials and Methods in sufficient detail to enable researchers in other laboratories to reproduce the experiments described in the manuscript.
- Use the generally accepted nomenclature in their field; define all non-standard terms.
- Report how often each experiment was replicated and, where applicable, whether the results were substantiated by replication over a range of conditions. Provide sufficient information to allow readers to distinguish between independent data points and technical replicates.
- Where applicable, state whether the samples were randomized and, if so, the method of randomization.
- Where applicable, state whether the experiments were blinded and, if so, how.
- Where applicable, specify whether the authors undertook a sample-size estimate when planning the study and, if so, how.
In manuscripts that report the development of novel chemical tools, or the use of known compounds, the authors should use systemic nomenclature to refer to the compounds, based on the guidelines developed by the International Union of Pure and Applied Chemistry (IUPAC); informal common names, such as dioleoylphosphatidylcholine, cholesterol or sucrose, are also acceptable. In the case of novel chemical tools, the authors are required to provide the exact structures of the compounds (prepared according to the IUPAC recommendations) and are encouraged to explain how each reported chemical structure was identified according to established standards in the field.
3. DATA AND IMAGE PROCESSING
Authors should make every attempt to minimize the post-acquisition processing of data. Some degree of processing may be unavoidable and is permitted provided that the processing procedure is described and the final data accurately reflect the original. In the case of image processing, alterations must be applied to the entire image (e.g., brightness, contrast, color balance). In case this is not possible (e.g., alterations to a single color channel on a microscopy image), any alterations must be clearly stated in the figure legend and in the Materials and Methods section. Groupings and consolidation of data (e.g., cropping of images or removal of lanes from gels and blots) must be explicitly indicated in the appropriate figure and figure legends.
Data comparisons should only be made among comparative experiments, and individual data should not be reused across multiple figures. When data are used multiple times (e.g., multiple experiments were performed simultaneously with a single control experiment), this must be clearly stated in each figure legend.
If it is deemed necessary for proper evaluation of the manuscript, authors will be required to make the original unprocessed data available to the handling editor. The figures in all accepted manuscripts will be examined before publication.
4. MATERIALS AND DATA AVAILABILITY
One of the terms and conditions of publishing in BJ is that authors agree to distribute materials (cells, DNA, antibodies, reagents, organisms and software) and protocols developed during the course of published studies, as well as the data that were generated, to qualified researchers for their own use.
Materials and protocols should be made available with minimal restrictions and in a timely manner, but it is acceptable to request reasonable payment to cover the cost of maintenance and transport of materials. Exceptions to this would include cases where supplies of purified proteins or reagents are highly limited, or where the transfer would violate compliance considerations. Other exceptions will be considered on a case-by-case basis by the Editors. Whenever feasible, data sets should be made available to readers from the date of publication, and must be provided to the handling editor upon request for the purpose of evaluating the manuscript.
Articles in BJ are based on many different kinds of data, and community-endorsed data standards and public repositories are not available for all types of data sets. When there are community-endorsed standards and repositories, the full data set must be submitted to an appropriate repository, and accession numbers must be provided in the article.
Examples of data sets and public repositories are listed below (The list of repositories is not intended to be complete and authors may deposit their data in public repositories not listed below, as long they state where the data are deposited).
- STRUCTURES OF BIOLOGICAL MACROMOLECULES
The atomic coordinates and related experimental data (structure factor amplitudes/intensities and/or NMR restraints) must be deposited at a member site of the Worldwide Protein Data Bank (wwPDB). Authors must agree to release the atomic coordinates and experimental data when the associated article is published.
For manuscripts reporting electron cryomicroscopy (cryo-EM) density maps and fitted coordinates, authors are encouraged to include with their submission either maps and coordinates, or a movie that shows an interactive session describing the map and the fit of the model in sufficient detail. Authors must agree to make the images and relevant metadata needed to reproduce a published EM reconstruction available when the article is published, preferably by depositing in the Electron Microscopy Pilot Image Archive (EMPIAR) or upon reasonable request. Cryo-EM-derived density maps must be deposited into the EMDataBank (EMDB), through one of the partner sites (EM Data Bank in Europe or EMDataBank). Atomic coordinates fitted to EM maps must also be deposited to a wwPDB member site. If an article discusses a protein structure only at the level of the main chain alpha carbon atoms, only alpha carbon coordinates need be deposited. If the discussion involves higher‐resolution data, a full coordinate list must be deposited.
BJ encourages authors to deposit the information before the manuscript is submitted because the corresponding database IDs must be included in the manuscript, and the handling editor should be informed no later than the completion of the editorial process that the necessary information has been sent to the appropriate data bank. BJ recommends that authors include the PDB validation report as part of the online Supporting Material for all new submissions describing results of X-ray and NMR structure determination.
For articles that include NMR spectra, authors must deposit the chemical shifts, coupling constants, relaxation data and other relevant data in the Biological Magnetic Resonance Data Bank (BMRDB) prior to submission of the manuscript and include the BMRDB-ID in the manuscript.
For articles that include circular dichroism spectra of proteins, authors must deposit the raw, processed, and metadata in the Protein Circular Dichroism Data Bank (PCDDB) prior to submission of the manuscript and to include the PCDDB-ID in the manuscript.
- SMALL MOLECULE STRUCTURES
The atomic coordinates and related experimental data (structure factor amplitudes/intensities and/or NMR restraints) of small molecules (lipids, amino acids, and other biologically active small molecules including drugs) must be deposited in the database maintained by the Cambridge Crystallographic Data Centre (CCDC). Authors must agree to release the atomic coordinates and experimental data when the associated article is published.
- DNA AND PROTEIN SEQUENCES
Authors must deposit their sequence data in the appropriate database among the following:
- Protein sequences: UniProt
- DNA and RNA sequences: GenBank, European Nucleotide Archive (ENA), DDBJ, Protein Data Bank, UniProt.
- DNA sequencing data (traces and short reads): NCBI Trace and Short-Read Archive, ENA’s Sequence Read Archive.
- Deep sequencing data: deposit in GEO or ArrayExpress upon submission to the journal.
- The sequences of all RNAi, antisense, and morpholino probes must be included in the article or deposited in a public database with the accession number provided in the article.
- Human genomic data containing newly described SNPs and CNVs identified in control samples must be deposited in an appropriate repository such as dbSNP, the Database of Genomic Variants Archive (DGVa) or the Database of Genomic Structural Variation (dbVAR).
Articles based on genomic data generated from HeLa cells must comply with the NIH HeLa Genome Data Use Agreement, by affirming that the NIH has approved use of the data and acknowledging the contributions of Henrietta Lacks and her family to the research.
- PROTEOMICS AND PROTEIN INTERACTIONS DATA
Authors must deposit their data in the appropriate database among the following:
- CHEMICAL COMPOUND SCREENING
The PubChem databse should be used.
- COMPUTATIONAL SYSTEMS AND PHYSIOLOGICAL MODELS
SBML and CELLML are two emerging standards for encoding computational models related to systems biology and physiology. To assure public access to such computational models, authors should, where applicable, deposit their models in the CellML Model Repository or the Biomodels Database. Other public databases for models developed in NEURON or VCell should also be utilized where applicable. When this is not possible (e.g., in the case of MatLab or other code), authors should include their model code as a file in the online Supporting Material.
- CELL IMAGING DATA
Currently, there is no community-endorsed repository of cell imaging data. If practical, raw image data should be made available as supporting material. Also, authors should, where applicable, make the image processing code available, preferably as a plug-in to open source software such as ImageJ or Python.
Authors must declare the location and accessibility of any custom code and software central to the main claims of their paper. We recommend deposition of source code GitHub together with a listing on Zenodo, which will assign a Digital Object Identifier (DOI) to make the upload uniquely citeable; this DOI should be reported in the manuscript.
The standards for the public deposition of data continue to evolve, and the above list will be revised at regular intervals. If there is no community-agreed-upon format (such as for imaging, molecular dynamics, electrophysiology, and single-molecule studies), if there is no public repository, or if the data sets are too large to submit to the journal online, authors are encouraged to make the data available on request.
5. CONSIDERATION OF REFUTATIONS
It is the policy of BJ that, once it has published an article, it accepts responsibility to consider for publication manuscripts that refute the interpretations of the results in the original article according to the Journal’s established standards for evaluating the suitability.
Updated July 2015