Ever heard of the reproducibility crisis, the four horsemen of irreproducibility, or do you know what a pre-registered report is? The answers are irrelevant! Like climate change, nothing is more important than this…
We will venture together to a place many have spoken of, but few have gone to – trying to reproduce the results of scientific papers . After a quick introduction into reproducibility and the setup of the hackathon, the participants will be split into teams and assigned publications to reproduce. Teams will collaboratively find ways to reproduce, to gain access, and to discover new sources of tolerance, navigating through the (potential) hurdles of replicating papers. Each team will present their findings, detailing their journey of attempting to replicate the results in a final discussion round. This will include insights gained, obstacles encountered, and suggestions for improving reproducibility.
Please note: This is not an introduction to programming or data analysis. Participants should have a favorite programming language and skills in analyzing data on their own. Participants should bring their own laptops (incl. power cords) and can suggest publications for reproduction (but do not have to).
By joining the SIB Days, you agree to abide by the SIB's Code of Conduct
We’ll track of discussions and collaborative notes on the event hackpad
As all ReproHack events, we strive to make this event open and inclusive to all. As such the event is governed by the ReproHack Code of Conduct.Please read it before participating. By participating, you are expected to uphold this code.
We’ll start with a brief review of the available papers and then form into teams. Feel free to work on your own if you prefer but we highly recommend you discuss your experiences with fellow participants as you work.
Add your details in the participants section of the hackpad so we have a complete list of participants (e.g name, affiliation, etc).
In your teams, (or individually if you prefer) decide which paper you wish to work on.
Create user accounts on the Hub for all reviewers that want to be associated with the review.
Register your team and paper by logging the title of the selected paper and the name(s) of the reviewer(s) in the hackpad. You can use the following template:
### **Paper:** <Title of the paper reproduced>
**Reviewers:** Reviewer 1, Reviewer 2 etc.
Follow any instructions/documentation associated with the papers and try and reproduce the work. As you work through your paper, keep in mind the main points on which feedback to the authors will provided, Access, Reproducibility, Documentation and Reusability (see our participant guidelines for more information). It might help to have a look at the Review form before you begin and keep notes during your review. Feel free to use the event hackpad to record general findings you wish to share with the group.
We’ll come together during the day to discuss progress and troubleshoot any sticking points.
Should you finish reproducing your paper quickly, feel free to explore the work more deeply. For example, you could try and run additional analyses, create new plots or even combine materials with your own or other open materials available on the web!
Should you produce any additional materials relating to your reproduction during the session (e.g. a markdown report, jupyter notebook, issue or pull requets in authors repository), feel free to share it publicly and add any links to such materials to the hackpad.
The most important part of the day is recording our experiences as feedback to the authors. Please make sure to complete a Review feedback form for the paper you've selected, ideally, by the end of the day. Please also remember to be kind and constructive. Reproducibilty is hard and all authors submitting their papers have been incredibly brave. Feel free to browse any public feedback submissions to get inspiration.
You can use the hackpad to take notes and summarise your experiences.
Feel free to contribute any additional thoughts on our collaborative hackpad. These can help form the basis for a blogpost on the event.
There is a numerical benchmark reported in Fig. 4 with absolute runtimes and memory usages that can directly be reproduced with the provided source code. The benchmark was performed on the author's computer, and since numerical performance and parallel scaling can be somewhat hardware-dependent, it would be of interest to see whether a performance that is comparable to the one reported in the paper can be reproduced by others on their own computers in practice. The benchmark simulates a growing tissue from one to 10,000 cells in just ten minutes, so this offers an easy entry point into tissue modeling and simulation. No input data is needed to reproduce the output. The program has no dependencies.
The methods are widely applicable to other DNA sequence clustering problems. Someone may obtain contradicting results with a new algorithm. In such a case, rerunning our scripts on the same or new data may help elucidate the source of the differences between the results.
This is a seminal paper on reproducibility in cancer biology. It should be a gold standard for reproducible research work. Therefore, it should be attempted to reproduce it. Supposedly, this will be pretty easy to reproduce and can be used as a *positive control* in repro hacks!
We are in the process of improving the documentation of the tools used in the analysis of wastewater. New upcoming website: https://dryak.github.io/vpipe/ Draft of the upcoming tutorial: https://gist.github.com/DrYak/e28d5d523e644ea455748d540a32ad4d Currently latest version of V-pipe: 3.0.0.pre1