mroswell / vaxscience Goto Github PK

Complete this tutorial:
https://www.nlm.nih.gov/bsd/disted/meshtutorial/introduction/index.html

Widen the Abstract and Affiliation fields

In the table, replace the display of the list with semicolons

 {{ cell.value | replace('", "','; ') | replace('[\"','') | replace('\"]','')}}

from https://github.com/mroswell/list-N/blob/9a723ceeb2cb56c136dda626ef13382e0f6d8986/templates/_table.html

Dedupe affiliations

How to convert

Purpura, Thrombocytopenic, Idiopathic/*chemically induced
Purpura, Thrombocytopenic, Idiopathic/*complications/pathology/therapy
Purpura, Thrombocytopenic, Idiopathic/blood/*chemically induced/immunology/therapy
Purpura, Thrombocytopenic, Idiopathic/blood/*etiology
Purpura, Thrombocytopenic, Idiopathic/blood/*etiology/immunology
Purpura, Thrombocytopenic, Idiopathic/blood/*etiology/therapy

to just "Purpura, Thrombocytopenic, Idiopathic"

Somewhere in here:

            row['Main_Topic'] = [item for item in mesh if '*' in item]

enable updating the Spreadsheet to update the website

theme the website to React19.org header

Add permissions on New Sheet

Remove link field. Replace with linked ID fields on PMID and PMCID. Move PMID and PMCID over to the left.

PMID PMCID

so that url will be ../research/articles.. instead of ../articles/articles...

Parallel to #1

Run transform.py on articles.json as follows

% brew install jq
% python3 transform.py | jq --monochrome-output . >articles2.transformed.json

Note that in this sample record, one of the Demographic entries ( "*Peripheral Nervous System Diseases") didn't move over from the MeSH field to the Main_Topic field. In some records they all properly move over, but sometimes they don't. There's seemingly no consistently about which items are missed.

      "MeSH": [
        "*COVID-19/prevention & control",
        "COVID-19 Vaccines/adverse effects",
        "*Peripheral Nervous System Diseases",
        "SARS-CoV-2"
      ],
      "Main_Topic": [
        "*COVID-19"
      ],

Run transform.py on articles.json as follows

% brew install jq
% python3 transform.py | jq --monochrome-output . >articles2.transformed.json

Note that in this sample record ("PMID": 34668274), two of the Demographic entries ("Aged" and "Middle Aged") didn't move over from the MeSH field to the Demographics field. In some records they all properly move over, but sometimes they don't. There's seemingly no consistently about which items are missed.

      "MeSH": [
        "2019-nCoV Vaccine mRNA-1273",
        "Aged",
        "BNT162 Vaccine",
        "*COVID-19",
        "COVID-19 Vaccines",
        "*Guillain-Barre Syndrome",
        "Middle Aged",
        "*Peripheral Nervous System Diseases",
        "SARS-CoV-2",
        "Vaccination/adverse effects"
      ],
      "Demographics": [
        "Adult",
        "Aged, 80 and over",
        "Female",
        "Humans",
        "Young Adult"
      ]

The goal is that each item in MeSH that appears in demographic_list should move to Demographics (be removed from MeSH and appended to Demographics).

Once we figure this out, I'll be exploring more ways to break up the MeSH terms into more manageable fields.

By default, Datasette is offering these suggested facets in the following order.

Main_Topic, Title, Link(s), Demographics, Abstract, MeSH, Author(s), Affiliation, PMID, PMCID, PubDate, article_title, Section, Subsection, PubType, PubDate (date), Main_Topic (array), Demographics (array), MeSH (array), Author(s) (array), Affiliation (array)

Replace this with
Main_Topic*, Demographics*, MeSH*, Author(s)*, Section, Subsection, PubType

• Asterisked fields (the first four facets) should be displayed in the Array version. Remove the asterisks before displaying.

there may be some model code (though by a relative newbie coder) here:
https://github.com/mroswell/list-N/tree/main/plugins
or
https://datasette.io/plugins

Create subdomain: research.react19.org