I’ve posted about publication lag times previously. The “lag” refers to the time from submitting a paper and it appearing in a journal.

Publication lag times are still a frustration for researchers. Although preprints circumvent the delay in sharing science with others, publication is still king when it comes to evaluation. Contracts are short and publication delays can be long…

I recently saw a post comparing median publication lag times for genetics journals. This motivated me to update my code and rerun the analysis for cell biology journals to see what, if anything, has changed.

Methodology

I wrote an R package PubMedLagR which uses {rentrez} to retrieve the publication data from PubMed. Once we have this data, it is a matter of producing some plots which I have covered previously. To ensure that we are only looking at research papers, we use "journal article"[pt] as a search term, and also remove from the data anything else (Reviews, Commentaries etc.). Feel free to use it to look at other journals.

Caveats

Before we get started, there are some caveats. The analysis is only as good as the PubMed data. Not all journals submit their date information to PubMed, while for others it is incomplete (as we’ll see below). There are inaccuracies for sure. I found a paper that was supposedly submitted on 1970-01-01, more than 40 years before the journal started. Also, it’s well known that some journals “restart the clock” on a paper by rejecting it and allowing resubmission, then only counting the resubmitted version. So, comparison between journals is a little tricky, but it allows us to look at trends.

Let’s dive in

We can use the following code to grab the data we are interested in.

library(PubMedLagR)
jrnl_list <- c("J Cell Sci", "Mol Biol Cell", "J Cell Biol", "Nat Cell Biol",
               "EMBO J", "Biochem J", "Dev Cell", "FASEB J", "J Biol Chem",
               "Cells", "Front Cell Dev Biol", "Nature Communications",
               "Cell Reports", "Mol Cell", "Autophagy", "Cell Death Differ",
               "Cell Death Dis", "Cell Res", "Sci Adv", "Cell")
yrs <- 2006:2026
retrieve_journal_year_records(jrnl_list, yrs, batch_size = 250)
pprs <- pubmed_xmls_to_df()

The list of journals is somewhat arbitrary. I have included Nature Communications and Science Advances although they carry many other papers besides cell biology. I included Cell (even though there’s not much cell biology in there these days) and left out Nature and Science.

How many papers are in each of these journals and how has that changed over time?

There’s a huge increase in the number of papers published by Nature Communications, Science Advances and Cell Reports. Nature Communications is a behemoth, publishing ~12,500 papers in 2025.

There was a boom and bust in publications at Front Cell Dev Biol and Cells which could be due to reputational problems (like this). Other journals have declined. Most noticeably J Biol Chem, but others have taken a hit. The reasons behind these dynamics are discussed elsewhere.

Median publication lag time

We’ll use received-to-published as our measure of publication lag time. This is the time from submission to it appearing in the journal “in print”. It’s measured here in days.

For the last 5 years, the median lag time at Nature Cell Biology is over one year. Other outlets are close to this, e.g. Cell, Dev Cell; whereas others linger around 200 days, or lower in the case of J Cell Sci, FASEB J et al. The journal Autophagy is missing here because there are no data for it. Others, like Sci Adv, MBoC have only minimal data available. The shortest lag times were for Cells (which might not surprise some people).

The lion’s share of this lag time is the time from submission to acceptance (received-accepted), with a small contribution from the time taken to formally publish the article (accepted-published).

This is using the median time. Obviously, some papers whizz straight in, whereas others… don’t. Let’s have a look.

You can see many examples of papers that have lag times of 1, 2 or 3 years. I scaled all the plots to a maximum lag time of 3 years. There were several examples of papers with lag times of up to 7 years that looked to be genuine, but they distorted the view.

So what trends can we see?

The lag times are creeping up at some journals, but not at others. It’s possible to see some very short lag times (in amongst the longer ones) recently at Dev Cell, Mol Cell and EMBO J. I assume these are transfers in to the journal, leading to rapid publication. Although Cell also has many of those too, so perhaps there are other explanations.

To click though in more detail, I made some graphics for each journal which use ridgelines to look at the profile of lag times for each year.

And finally

Incidentally, nine of the top ten papers with the longest lag times were published in Nature Communications. The longest was this one (3263 days). Almost nine years! All papers have their battle stories and I’m sure this one has a tale to tell.

—

The post title comes from “Hold On Hope” by Guided by Voices.