The 23rd annual Nucleic Acids Research Web Server Issue 2025

Welcome to the 23rd Web Server Issue of Nucleic Acids Research with 72 web servers from and for different scientific and medical fields.The trend toward using large-language models (LLMs) has reached scientific web servers, and consequently, we received a number of proposals describing LLM-based web servers.Surprisingly, none of those that employ an LLM to interact with the user made it into this Issue.Some of them were already filtered out in the first step, e.g. because they did not give better answers than existing general-purpose LLMs or came without any documentation at all-refusing even to provide it when prompted.Others did not pass our thorough in-house technical review or external expert evaluations, often due to unreliable output or "hallucinations."Chatbot-style interfaces also demand comprehensive documentation-not only for the interaction mechanisms but also for the underlying data and, most importantly, the limitations of the system.This includes clearly outlining the circumstances under which the model might generate incorrect or misleading results.Another trend is the use of cloud-based solutions.These pose a great challenge for the Web Server Issue because they usually require users to log in-and that is something that we do not allow (unless for sensitive data such as human medical or genetic data).An additional concern is the potential need for payment or at least credit card registration when using commercial cloud platforms, which conflicts with our fundamental principle of providing free and open-access web servers.We do, on the other hand, recognize that the provision of IT infrastructure does not come for free and that many groups are simply not able to cover the cost for high-performance computing or large data storage from their own budgets.While our yearly editors' meeting takes place after this editorial's deadline, I encourage all future submitters to review our updated submission guidelines at https: // academic.oup.com/nar/ pages/ Submission _ Webserver , which will be revised by August.You may also notice a lower number of single-cell analysis web servers in this year's Issue.One reason is that many of the suggested web servers were too specialized (e.g.aimed at only one specific tissue) or did not provide clear advantages over existing solutions.However, a main problem here was that many were web versions of existing R packages and would require to use R to prepare the input data and sometimes even for further downstream analysis, limiting accessibility for nonbioinformaticians.One clear aim of the Web Server Issue is to facilitate data analysis or lab tasks for researchers who do not want to use the command line.A web server is of limited use if it demands extensive pre-processing before it can be used or if it only provides plain text files as output, but does not provide contextual information (e.g.gene annotation data or external links) and options for filtering and sorting the results.