Methods section too short? Use online protocols to make complex techniques understandable

The human brain is beautiful and complex. When viewed with the proper tools, the seemingly homogeneous mass is revealed to have intricate layers, folds and structures. In Sten Linnarsson’s laboratory at the Karolinska Institute in Stockholm, where I did my graduate research, we developed methods to uncover these patterns on the basis of differences in gene expression from cell to cell. We used this information to identify and distinguish different cell types in the brain and to understand their spatial organization.

Our newest method, enhanced electric fluorescent in situ hybridization (EEL FISH)¹, builds on existing methods to detect the expression levels of hundreds of genes — but with a trick. EEL FISH uses an electric field to drive RNA from a 3D tissue section onto a flat surface, so that all the RNA molecules are in the same imaging plane — a modern twist on the northern blot, a classic technique used to detect RNA sequences. This method accelerates tissue imaging, and thus can cover a larger area without prolonging the experiment. Alejandro Mossi Albiach, a graduate student at the Karolinska Institute, and I have used the tool to reveal patterns in the developing human brain² and the complex cellular organization of the adult human brain³.

NatureTech

EEL FISH required intricate instrumentation and years of development by a large team of researchers. Sharing the protocol is our duty and an essential part of the scientific process. It allows others to replicate and extend the method, and to scrutinize our results. But for a tool as complicated as EEL FISH, conveying all the pertinent details is difficult. Here’s how we overcame that limitation.

Go step by step

For any technique, the methods section in the published article should contain all the information needed to replicate the results. But the level of detail in this section can vary considerably. Authors often provide only brief summaries with minimal detail. For most studies — and especially for new methods — it is better to describe not only what was done, but how.

Step-by-step protocols provide a version of the procedure that is easier to follow and implement than what is described in a standard methods section. Such protocols usually contain a detailed list of materials with part or product numbers and troubleshooting guides, and can be written for both wet-lab and computational protocols — giving new users a good place to start. Some protocols, such as those published on Nature Protocol Exchange (published by Springer Nature, which also publishes Nature), Cell Press’s STAR Protocols, PLOS Lab Protocols and Bio-protocol, are hosted by journal publishers, but there are other options, including Protocols.io and OpenWetWare.org.

We published two EEL FISH protocols on Protocols.io, one for the wet-lab part of the method and one for building the necessary hardware, which together have received thousands of views.

Linked in the published method, these are great additions to the scientific record because they not only contain detailed information on the technique, but also offer a forum for questions and comments, which everyone can see. For a related method that we published in 2018, we received more than 30 requests for clarifications, explanations or alternative reagents. These questions helped us to create a more comprehensive EEL FISH protocol.

Methods on Protocols.io can also be updated post-publication. For instance, one of the reagents we used in EEL FISH was no longer commercially available by the time we published our protocol, so we provided a recipe for a working replacement in a comment. This kind of post-publication editing makes it easier to support methods over the long term.

Communicate common pitfalls

Whatever the level of detail in the publication, methods rarely work correctly on the first try. But in our experience, most issues that first-time users run into aren’t new; developers encounter them first, during development. So, don’t be shy — ask for help.

Authors are usually happy to assist researchers who want to implement their method (provided the users first try to fix the problem themselves and are polite). It might even be possible to visit the lab that developed the tool to learn about it directly.

Keep an eye out for common mistakes by paying close attention to what goes wrong when users try the method for the first time. In EEL FISH, for instance, we mount samples on thin pieces of glass, which are fragile and need some skill to handle. I learnt this the hard way at the beginning, but later forgot as I got used to working with the glass; watching a new user experience the same frustration brought that knowledge back to me. Now, I recommend that beginners put on safety goggles and break a few glasses to learn their physical limits before attempting the technique. That advice usually gets me funny looks, but it’s saved many experiments.

You can also use pictures or video to document key steps. In our guide to the EEL FISH hardware, we included images of almost every step so that even people with little engineering experience could build the system. We also like to include schematic summaries, such as flow charts for experimental pipelines and data-analysis steps, to help the reader to visualize the procedures. Alternatively, video journals such as the Journal of Visualized Experiments can be a great way to disseminate methods or get started with using them.

Define the ‘obvious’

Methods are usually written for an audience that already understands common practices in the field. This can create difficulties for readers who have a different background, or who are reading the article long after publication. In many cases, what was common knowledge at the time of publication no longer is.

EEL FISH, for instance, shares many features with DNA microarray fabrication and processing, which was popular in the 1980s and 1990s but has since gone out of fashion. Without the knowledge about microarrays that would have been common decades ago, I had a hard time piecing together fabrication procedures from published papers. Difficult steps were usually described in detail, but routine steps, such as washes or incubations — and the type of equipment used to accomplish them — were not described at all. We try to be mindful of this in our own protocols and to describe even ‘obvious’ steps in detail.

If possible, have a colleague or student test-drive the protocol before publication. We were lucky that a new student started in the lab while we were preparing the EEL FISH manuscript, and they could test the protocol to see whether it was clear. Their experience highlighted places where we were relying too much on experience and not fully describing the procedure. You can also try feeding the steps to a chatbot: in January, bioengineer Arjun Raj at the University of Pennsylvania in Philadelphia demonstrated that ChatGPT can identify ambiguities in written protocols.

When it comes to experimental methods, the devil is in the details. By outlining those details clearly and comprehensively, we can make methods more accessible, transparent and engaging — not to mention, useful.

Competing Interests

L.E.B. is a shareholder in EEL Transcriptomics AB, which owns the intellectual property of the EEL FISH method (“Spatial RNA localization method”, provisional US patent application 63/139,701), and is an unpaid ambassador for Protocols.io.