TEMAS

Tutorial

Things to know for all steps

  • If there is no pop-up on your screen, it means that you have something to do. You may have to scroll down to show the actions to be performed.
  • When you click on the Download buttons, if a choice is offered to you (with Firefox web browser for example) prefer "save" option than "open" option.
  • When you click on the Download buttons, files are automatically downloaded to your browser's default folder. If you want to change this default folder you can follow this link

TEMAS Step 1: Global Search

  1. Define search parameters
    • Enter your search keywords
      • You can use advanced search since it respects PubMed syntax.
      • Use parentheses wisely
      • The use of the "NOT" operator is not recommended.
      • For "phrase searching" (e.g. breast cancer) use "_" instead of enclosing the phrase in double quotes (e.g. breast_cancer instead of "breast cancer")

      Example:
      (breast_cancer[MeSH Terms] OR breast_cancer[Title/Abstract]) AND (risk_factor[Title/Abstract] OR risk_factor[MeSH Terms])

    • Select the date range for the articles you want to retrieve (default setting is a search over a range of 10 years from today's date)
    • Click on the "Save search parameters" button and wait until "Done!" appears on your screen
  2. Create database
    • Select the maximum number of articles you want to retrieve (default setting is the number of PubMed answers)
    • Click on the "Create database" button and wait until "Done!" appears on your screen
    • A preview of your database is displayed on the right of your screen
  3. Download database
    • Click on the "Download database" button to save the formatted database.
  4. Go to Step 2
    • Click on the "Step 2 : First sort" link to go to Step 2

TEMAS Step 2: First sort (removing abstact containing off-topic words)

  1. Import data
    • In the left sidebar, upload the database you saved in step 1 by clicking on the "Browse..." button
    • Info! Default options to import your database are already selected, but if you import a database not formated in step 1 please select the right options.

    • When "Upload complete" is displayed, save the database clicking on the "Save database" button. "Done!" should appear on your screen
  2. Removing off-topic abstracts
    • With the slider, select the number of off-topic words you want to eliminate, and enter these words in the boxes that appear
    • Info! It is advisable to use stems of words like "metastas" instead of "metastasis" and "metastases".

  3. Management of articles without abstract
    • Check the box below to remove PubMed results without abstracts (strongly recommended)
  4. Sort Database
    • Click on the "Sort database" button and wait until "Done" appears on your screen
  5. Sort results and download
    • The number of articles before and after sorting are now displayed
    • Click on the "Download sorted database" button to save the sorted database.
  6. Go to Step 3
    • Click on the "Step 3 : Clustering exploration" link to go to Step 3

TEMAS Step 3: Clustering exploration

  1. Import data
    • In the left sidebar, upload the database you saved in step 2
    • Wait until "Upload complete" is displayed
    • Save the database clicking on the "Save data" button. "Done!" should appear on your screen
  2. Classification
    • Click on the "Compute classification" button, "Computation in progress..." should appear on your screen
    • This is the longest computational step, please wait until "Done!" appears on your screen
  3. Classification parameters
    • "Number of clusters" is the most important parameter. Select a number of clusters allowing fairly good discrimination between classes without unnecessarily multiplying classes
    • "Max number of terms to display" indicates the number of terms displayed in each class on the graph
    • "Text size" allows you to change the size of the text on the graph
    • Click on the "Download graph" button if you want to download the graph once you have entered the parameters
    • Click on the "Valid number of clusters to continue" button
  4. Class and term selection
    • Select the class that contains the terms that seem relevant to you. The class numbers are indicated under the barplot on the left of the graph.
    • At this stage, you have 2 possibilities :
      • Single-word extraction (e.g. breastfeeding)
        • Select in the list of words the term(s) that interest you
        • You can type the beginning or part of the word to display all the terms that contain the letters you entered
        • You can make a multiple selection (e.g. premenopause postmenopause). All the articles which contain at least one of the terms entered will be selected
      • Multiple-words extraction (e.g. oral contraceptive)
        • Select in the list of words the main term that interest you (e.g. contraceptive for "oral contraceptive")
        • You can type the beginning or part of the word to display all the terms that contain the letters you entered
        • Enter a complementary word associated to your main word (e.g. oral for "oral contraceptive")
        • Select the maximum distance from the main word (in number of characters) you allow to look for the complementary word
    • Click on the "Extraction" button
    • A preview of your database is displayed under the graph
  5. Download extracted database
    • The number of articles extracted is displayed
    • Click on the "Download data" button to save database.
    • Warning! Please don't change the beginning of the file name (keep at least "shiny.classif_your-selected-word_" ).

  6. Go to Step 4
    • Click on the "Step 4: Final Sort" link to go to Step 4

TEMAS Step 4: Final sort

  1. Import data
    • In the left sidebar, upload the database you saved in step 3
    • Info! Default options to import your database are already selected, but if you import a database not formated in step 3 please select the right options.

    • When upload is complete, save the database clicking on the "Save data" button, and wait until "Done!" appears on your screen
  2. Sorting choice
    • There are 3 choices for sorting:
      • Preset
        • This choice will keep all the articles that contain at least one of the following terms "OR", "RR", "HR", "relative risk(s)", "odd(s)", "hazard ratio(s)
      • Custom
        • This choice will keep all the articles that contain at least one of the words that you will write in the boxes that appear (up to 6 words)
        • Warning! Custom sorting is case sensitive! If you enter "Survival" the search will not keep articles which only have the word "survival" in lowercase. To avoid this problem you can either enter "urvival" or enter both "Survival" and "survival"

      • Both = Preset + Custom
        • This choice will keep all the articles that contain at least one of the following terms: "OR", "RR", "HR", "relative risk", "Odd(s)", "hazard ratio(s) or one of the words that you will write in the boxes that appear
    • When your choice is done, click on the "Sort" button, and wait until "Done!" appears on your screen
  3. Download final data and see these articles on PubMed
    • Click on the "Download final data" button to save this final database.
    • You can also click on the "PubMed link" to view (in a new tab) these articles directly on PubMed.

TEMAS Step Bonus: Automated database merging

  1. Import data
    • TODO
  2. Merge databases
    • Click on the "Merge databases" button
  3. Merging results
    • The number of articles before and after merging and removing duplicates are now displayed
    • Click on the "Download merged database" button to save the merged database.

Thanks for using TEMAS

Developed by Emmanuel Bonnet

References

  • Julien Barnier (2020). rainette: The Reinert Method for Textual Data Clustering. R package version 0.1.1.
    CRAN.R-project.org/package=rainette
  • Winter, D. J. (2017) rentrez: an R package for the NCBI eUtils API The R Journal 9(2):520-526
  • Winston Chang, Joe Cheng, JJ Allaire, Yihui Xie and Jonathan McPherson (2020). shiny: Web Application Framework for R. R package version 1.4.0.2. https://CRAN.R-project.org/package=shiny
  • Reinert M. Une méthode de classification descendante hiérarchique : application à l’analyse lexicale par contexte. Cahier de l’analyse des données. 1983;3(2):187-198.