How do you find science papers? Part 1: Tips for pros and pre-pros alike

In March, our laboratory entered a quiescent period to help slow the spread of coronavirus. We were fortunate enough to be allowed to resume our lab work again pretty quickly after a period of working from home. Though I can’t say I was (am?) at peak productivity during worldwide peril, I did still manage to nudge a writing project closer to the finish line while I was working from home. The paper I started during the lockdown and am finishing now is (I hope) going to be my first publication as a post-doc here at MUSC, and it’s given me a chance to reflect a little on the basics of scientific writing. 

Joining a new lab and progressing through training means that you are probably getting familiar with a whole new section of scientific literature. The need to demonstrate a good understanding of your novel field really comes to a head when you sit down to communicate your findings for publication. In the introduction, you must lead the reader swiftly and precisely from the general issues facing the field to the specific things you did to address those gaps in knowledge. In the discussion, you need to contextualize your findings elegantly within the most up-to-date research on your topic. It’s hard! And it’s especially hard to write a paper when you are new to a subject, as when you enter a new lab as a grad student or a post doc, or even if you are a lab head that is preparing to shoot for a grant a little outside your wheelhouse.

“Joining a new lab and progressing through training means that you are probably getting familiar with a whole new section of scientific literature. The need to demonstrate a good understanding of your novel field really comes to a head when you sit down to communicate your findings for publication.”

Anny Gano, postdoctoral fellow

As I am struggling to write my paper, I wanted to share with you both classic and innovative tools that make it easier to get your feet wet in a new pocket of science literature. In Part I of this two-part post, I will describe a classic tool that you probably learned to use in your college research methods class: PubMed. PubMed is so ubiquitous that it seems deceptively simple – everyone knows how to look up a specific paper in its search engine, right? Yet, an informal poll I conducted around my building revealed that 1/1 grad students and 2/2 post docs queried are not sure what the full scope of PubMed is and don’t think they are using it to its full potential. So, let’s talk Lit Search 101.

PubMed is a database of citations for scientific publications and does not automatically provide full text articles. When you search for a topic, about a third of what you get will be links to free full-text articles, and the rest will be abstracts for papers that are behind a paywall. However, if you are part of an institution like MUSC, you will be able to access most of these papers because of agreements that universities typically have with publishers. If you are affiliated with MUSC, you can use the PubMed link I provided (note that it has ‘musc’ in the URL) or access PubMed via our library website and, after logging in, you should be able to access a lot of papers without being asked for payment. If a full text is not available in the library’s collection, you can also request a free copy via the Interlibrary Loan service. Fun fact! Scientists love to share their findings. If you can’t access a paper, don’t be shy: google the first or last author’s contact info and email them. In my experience, he or she will gladly send you a pdf of the paper and may even be willing to answer questions about the work! Basically, you should never have to pay to access knowledge. 

So, how do you actually do a PubMed search? The search bar is versatile, and you can further narrow down results as you go on with the query. As a starting point, you can search snippets of information you may vaguely recall from a talk (like pieces of a paper title, one or two authors, maybe a general topic) and PubMed will use fuzzy logic to find the info for you even though you may have gotten a word or two wrong. 

If you are a frequent flyer, you probably want to get more sophisticated. The best way to hit Lit Search level 200 is to use PubMed’s filtering and fine-tuning options. If you are altogether new to this, you should probably start by learning how PubMed searches actually work. You can read the FAQs on the main page or do a tutorial, focusing on boolean operators, automated term mapping, MeSH terms and tags (or you can ask a librarian for virtual training – they’re awesome and can walk you through what you need to know!). Using these options can help you find exactly what you need, faster. For example, if you only remember a few last names of the people who did the kind of work you want to read about, you can tell PubMed that you specifically want to search for authors by adding the field tag “[au]” after the names. Oh, you want to find an Anny Gano first-author paper but, dear me, there are just too many papers with this Anny person as an author in general? Better make it “Gano [1au].” Knowing the field tag system will make your searches more precise.

It is also a good idea to learn about options PubMed provides for filtering your search results. You can restrict your search to a particular time period when the papers were published (use the slide bar on the left or ask for papers published “in last x months/years/whatever”), ask for specific article types such as  only review papers or just clinical trials, narrow down the research by species, and so on. These and other options can be found in the sidebar to the left of the search results. It is important to note that in the legacy version of PubMed, results were sorted by recency, whereas the new/current version will use an algorithm to give you “best fit” by default. If you don’t like this, you can change the sort order under “display options.”

If you are researching a general topic instead of looking up researchers whose names you already know, you can type key words into the search bar, as well. For example, if I am looking for information on alcohol use disorder and neuroimmunology, I may type in “alcohol immune” into the search bar and narrow down my results from there. Retrieved articles will show the search words bolded (and highlighted, if you are logged into your NCBI account) within a small snippet of the abstract. The neat part about PubMed is that it will take a few liberties and give you related overlapping terms. For example, if you search for “alcohol,” you might also get articles for “alcoholism,” or retrieve articles about “gastroesophageal reflux” based on your search term of “acid reflux.” If you want to search for specific phrases or snippets of phrases, you should read the PubMed FAQs and learn the exact settings for your needs.

Are you ready to take your relationship with PubMed to the next level and go steady? Create an account. If you do this, you can set up alerts so that when a new paper pops up that satisfies your saved search parameters, PubMed emails you about it. When you sit at your computer and use PubMed, it will store your search locally for eight hours. If you are doing searches while logged into your account, you can save a search and hit it again later. For example, if your awesome review paper gets stuck with reviewer #2 for an obscene amount of time and, when you get it back for revisions, the references are woefully out of date, you can go back to your saved search and grab any new papers that have come out in the meantime, giving you the freshest sources for your resubmission. Saving searches and signing up for email or RSS delivery are great ways to really get to know your new science – it creates a pipeline of relevant and timely literature that keeps you up to date on what’s going on in your field. As a new grad student, this is a great way to passively collect literature.

For those of us who have been using PubMed for a while, it may be interesting to note that the website has recently (well, November 2019) undergone its biggest makeover in a long time. Some main changes aren’t visually apparent but, as a result of updates to the programming, provide a faster and smoother search experience. There are also some features that are more noticeably altered. You may note that the advanced search options have undergone a drastic change – you now build the search one line at a time. Yet other features are entirely new, such as a forward-search option, which show you articles that have cited the paper that you are currently reading, allowing you to more easily expand your lit search surrounding a particular topic.

Fun fact! In 1986 the National Library of Medicine’s MEDLINE database of citations became available over the Internet for the first time via a new software. Because scientists know how to “rock” a good pun, this precursor to PubMed was named Grateful Med! PubMed remains a classic resource that is a cornerstone for any scientist’s virtual writing desk, but there are also new sleek writing tools cropping up that are revolutionizing lit searching. Check out my next blog post to modernize your lit search with brand new tools!

# # #

A special thanks to Christine Andresen from MUSC Libraries for an interesting seminar on PubMed updates and for taking the time to help me with this post.