Improved Search Now Available Across NCBI Databases


Earlier this year, we announced the release of a new and improved search feature that interprets plain language to give better results for common searches. This feature, originally developed in NCBI Labs and later released on the NCBI All Databases search, is now available across several NCBI resources: Nucleotide, Protein, Gene, Genome, and Assembly. Whether you are searching for a specific gene or for a whole genome, you will now retrieve NCBI’s best results regardless of the database  you search.

The image below shows the results for a search for human INS in the Nucleotide database. Even though this is a Nucleotide search, the results include relevant information from Gene, Protein, Taxonomy,  plus links to the NCBI reference sequences (RefSeq) as well as access to BLAST and the insulin gene region in NCBI’s genome browser, the Genome Data Viewer.KIS_nuccore_smallFigure 1.  The new natural language search result in the Nucleotide database from a search for human INS.

Try out this new search capability and let us know what you think. And keep visiting the NCBI Labs search page to try our latest experiments, which we’ll also announce here on NCBI Insights.

 

As-you-type-suggestions come to NCBI Labs


In a recent post, we described a new way to search our databases in NCBI Labs. We have now added a suggestions dropdown to the search bar that should make life easier for many NCBI visitors.

The as-you-type suggestions are simple, natural language-like queries we described in the previous post. They’ll help you avoid typos and save time if you’re searching for organisms with long or hard-to-spell names.

These suggestions are meant to direct you to high value results. As we improve the search experience, you may notice changes to the suggestions. We welcome your feedback on ways to enhance this new feature.

Here’s a quick look at what to expect:

dropdown suggestions ncbi labs search

Figure 1. As-you-type suggestions appear in a dropdown. Note how “human” is recognized as homo sapiens. Many common organisms are supported in this manner, e.g. “mouse”, “cow”.)

PubMed Journals will be shut down


Almost two years ago, we launched PubMed Journals, an NCBI Labs project. PubMed Journals helped people follow the latest biomedical literature by making it easier to find and follow journals, browse new articles, and included a Journal News Feed to track new arrivals news links, trending articles and important article updates.

PubMed Journals was a successful experiment. Since September 2016, nearly 20,000 people followed 10,453 distinct journals. Each customer followed 3 journals on average.

Though PubMed Journals will no longer exist as a separate entity, we hope to add its features into future NCBI products. We appreciate your feedback over the years that made PubMed Journals a productive test of new ideas.

NCBI Labs is NCBI’s product incubator for delivering new features and capabilities to NCBI end users.

Test drive a new sequence search experience at NCBI Labs


We know it’s not always easy to find the sequence data you’re after at NCBI. Maybe it’s because you’re no expert at constructing queries, and you end up with no results or too many results. Or maybe you’re an Entrez wizard, but creating a query full of Booleans and filters seems like overkill when you could just write a short natural language query, like you’re used to doing in Google.  The next time you search for a gene, transcript or genome assembly for a given organism, try the new search experience we’re piloting in NCBI Labs.

In NCBI Labs, you can now search for sequences using natural language and get the best results.

NCBI Labs transcript search interface

Figure 1. The new interface for specified transcript search.

The improved search experience now available in NCBI Labs addresses 3 types of queries that commonly fail in searches at NCBI: organism-gene (e.g. human BRCA1), organism-transcript (e.g. Mouse p53 transcripts) and organism-assembly (e.g. dog reference genome). For each of these query types in NCBI Labs, we now return NCBI’s highest quality sequence sets or reference and representative assemblies in an easy-to-view panel.

Example queries are shown below to get you started.

Continue reading

Try Our New, Experimental PubMed Search and User Interface in PubMed Labs


NLM needs your input. We are experimenting with a new PubMed search algorithm, as well as a modern, mobile-first user interface, and want to know what you think. You can try out these experimental elements at PubMed Labs, a website we created for the very purpose of giving potential new PubMed features a test drive and gathering user opinions.

Please note that PubMed Labs includes only a limited set of features at this time and not the full set of PubMed tools. The absence of a feature or tool on PubMed Labs does not mean we plan to eliminate it from PubMed; it simply means we are not testing it now!

The key elements we are testing are:

In PubMed Labs, search results are listed by best match. The best match button is at the top right; you can also click the "most recent" button (to the right of "best match") to sort by date.

  • A new search algorithm for ranking (ordering) the best matches to your query

Based on analysis of data obtained from anonymous PubMed search logs, we have developed a new algorithm that we believe does a much better job of sorting search results by their relevance, or “best match,” to your query. This new algorithm incorporates machine learning to re-rank the top articles returned.

We were so excited by results with this algorithm that we already implemented it in PubMed, but it is still experimental and we would very much appreciate hearing what you think. Part of our test in PubMed Labs is having best match be the default sort, instead of PubMed’s default of sorting by most recent articles. If you find that you prefer to sort by the most recent articles instead, it takes only a simple click of a button to do so.

Interested in specifics about the new algorithm? You can read more in this NLM Technical Bulletin.

Continue reading

PubMed Labs is now part of NCBI Labs


About two years ago, NCBI launched PubMed Labs, a gathering place for discovering and experimenting with new features and content for NCBI’s family of websites.  Over those years, we launched a few experiments that have helped us learn more about our customers and how we can serve them better.

Today we’re happy to announce that we’re expanding PubMed Labs to a broader set of experiments called NCBI Labs.

Why are we doing this?

  • To better convey the breadth of upcoming experiments on data, services, and websites that NCBI offers now and hopes to offer in the future.  You can expect to see new features, content, and other experiments from NCBI Labs in the coming months.
  • To reserve the name “PubMed Labs” for an exciting new set of experiments around biomedical literature and especially literature search.

What Will Change?

This blog’s menu item and blog category “PubMed Labs” will now appear as “NCBI Labs”. Existing links will continue to work.  We won’t be updating the old blog posts, for the most part, although some links on existing sites (e.g. on PubMed Journals) may be updated to use the new name.

Find, Browse and Follow Biomedical Literature with PubMed Journals


Following the latest biomedical literature can be a challenge, but NCBI’s new PubMed Journals will help you keep up-to-date.

PubMed Journals lets you:

  • Easily find and follow journals of interest
  • Browse new articles in your favorite journals
  • Keep up-to-date with a Journal News Feed containing new arrivals, news links, trending articles, and important article updates (retractions and more!)

The PubMed Journals home page.

Continue reading

PubMed Also-Viewed: Quickly find related articles


You’ve seen it before on shopping web site: you load a page displaying an item you want and see a list of other items that people bought with the one you’re viewing.

PubMed is free, but finding the important articles on a topic can cost a lot of time. To help you keep on top of the literature – with a little help from your fellow PubMed users – we are introducing a new type of link called “Articles frequently viewed together”. For some PubMed abstracts, you may see this link in the “Related Information” section in the right column.

PubMed Also-Viewed feature

Figure 1. The PubMed Also-Viewed feature.

Continue reading