HGEN 1988-2016. Interactive exploration of publication records

I used PubMed to find all the papers with HGEN affiliations and performed a few analysis on HGEN papers and authors. The results can be explored through an interactive application.

https://jmonlong.shinyapps.io/hgenpubs-app/

Here is a short story of these results but feel free to explore the results yourself.

Warning: There are some limitations and these numbers are as good as the PubMed records (see Limitations at the end).

HGEN first papers

The first HGEN paper seems to be a Biochemistry and Cell Biology paper from Ridgway et al. It simply showed that Dibutyryl cAMP inhibits expression of transformation-related properties in Kirsten murine sarcoma virus transformed Balb/c-3T3 cells despite continued presence of p21v-Ki-ras.

Out of the 6 papers out in 1988, three are from Charles Scriver.

HGEN growth phases

When did HGEN really started to expand ? Let’s look at the cumulative number of HGEN authors through the years. We can see three phases:

hgen-kickoff

  • 1988-2002 constant growth with around 19 new authors per year.
  • 2003-2009 much more people affiliated with around 104 new HGEN authors per year.
  • 2010-2016 a second increase with currently around 176 new authors per year.

Highlights

Using the number of citations from Scopus, we get an idea of the most influential papers for each year. This is what the first tab of the application is showing.

Hover over a point to see the title of the paper and find more information (e.g PubMed link) in the table below the graph.

giphy

Among the top stories, two methylenetetrahydrofolate reductase studies (1995 and 1998), a GWAS for type 2 diabetes (2007) or a BRCA1 characterization of breast cancer (2003).

HGEN author stats

Next I computed a few metrics for each HGEN author: number of HGEN publications, number of coauthors, number of mentees, year span, publication rate. This is what the interactive table in the second tab of the application is for.

You can search for people or reorder the different columns.

For example, we see that Dr Foulkes had the most publications with the department (99) and the most HGEN co-authors (323), but Dr Majewski had the most co-authors (798). Dr Rozen had the most mentees (30), defining a “mentee” as the first author on at least one paper with the mentor as last author.

Collaboration network

The department is pretty diverse, with HGENs in different locations and doing a wide variety of research. Can we spot scientific “friendship” as pairs of HGEN authors that publish a lot together ?

The third tab of the application shows the number of co-publications for every pair of HGEN authors. The fourth tab creates a network with the most fruitful collaborations.

Search the table for people or play with the network: scroll to zoom (names appear when you zoom enough) or move the nodes around (very satisfying).

For example, we can see a strong collaboration triangle between Drs. Mes-Masson, Provencher and Tonin, or some star-like topologies with Dr Rozen (or Drs Majewski or Foulkes) and some of their students.

hgen-network

Themes

Finally, we might wonder: what’s in all these papers ? As a proxy let’s parse the titles of the articles.

The fifth tab of the application creates a wordcloud with the titles. The sixth shows the usage particular words across the years.

Define the period, change the wordcloud parameters or search for specific words.

wordcloud

Here we see some expected words like “gene”, “genetic”, “syndrome”. We also see some recurrent words that are directly attributable to some researchers. For example methylenetetrahydrofolate reductase has been the theme of Dr Rozen’s lab and present in the wordcloud. A time analysis shows that this term is present almost every year in the title of a HGEN paper.

Limitations

There are several limitations to this small survey.

  • It is based on PubMed records which are not perfect. Sometimes all the affiliations are grouped into one and I might have counted some authors as HGEN when they are not. I tried to spot and filter suspiciously long affiliations. For this reason some publications were not counted either when they were real.
  • The number of citations per paper was extracted from Scopus. Different sources (e.g. Google Scholar) might give different results. Also publications before 1996 are not well tracked by Scopus.
  • I only take into account publications affiliated to the HGEN department. Some researchers don’t always use their HGEN affiliations, so many of their publications might not be counted.

Definitions

  • HGEN paper: with both “Human Genetics” and “McGill” in at least one of the affiliations.
  • HGEN author: author with “Human Genetics” and “McGill” in at least one of his affiliation in at least one HGEN paper.
  • Mentee/Mentor: A mentee is first author in at least one paper with his mentor. Both must be HGEN authors.
  • Publication rate: number of publications per year, after removing the first and last years they published (to be sure to count “full” active year).

 

— Analysis by Jean Monlong —

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s