Analysis of COVID-19 Clinical Trials: A Data-Driven, Ontology-Based, and Natural Language Processing Approach

Developed by Shray Alag, The Harker School, San Jose, CA

Publication       Coronavirus Report

References

Alag S (September, 2020). Analysis of COVID-19 clinical trials: A data-driven, ontology-based, and natural language processing approach PLOS ONE 15(9): e0239694. https://doi.org/10.1371/journal.pone.0239694 pmid:32997699 PLOS ONE PubMed/NCBI Google Scholar

PLOS ONE

Alag S (May, 2020). Unique insights from ClinicalTrials.gov by mining protein mutations and RSids in addition to applying the Human Phenotype Ontology PLOS ONE 15(5): e0233438. https://doi.org/10.1371/journal.pone.0233438 pmid:32459809 PLOS ONE PubMed/NCBI Google Scholar

PLOS ONE

SNPMinerTrials.com: Unique insights from ClinicalTrials.gov by mining protein mutations and RSids in addition to applying the Human Phenotype Ontology http://snpminertrials.com/ SNPMinerTrials

SNPMinerTrials
352,841

Total number of clincal trials.

4,038

COVID-19 related clinical trials

15,530

HPO nodes used in analysis

691

Number of unique MeSH terms associated with COVID-19 clinical trials

Reports

Data processed on September 26, 2020.

An HTML report was created for each of the unique drugs, MeSH, and HPO terms associated with COVID-19 clinical trials. Each report contains a list of either the drug, the MeSH terms, or the HPO terms. All of the terms in a category are displayed on the left-hand side of the report to enable easy navigation, and the reports contain a list of correlated drugs, MeSH, and HPO terms. Further, all reports contain the details of the clinical trials in which the term is referenced. Every clinical trial report shows the mapped HPO and MeSH terms, which are also hyperlinked. Related HPO terms, with their associated genes, protein mutations, and SNPs are also referenced in the report.

Drug Reports   MeSH Reports   HPO Reports  

Interventions

4,180 reports on interventions/drugs

MeSH

691 reports on MeSH terms

HPO

263 reports on HPO terms

Google Colab

Python example via Google Colab Notebook

API

JAVA APIs to access the data. Each Java class is a stand-alone program and does not require any other package beyond the Java core classes: Users can simply download a Java IDE, install Java, and run the class on that IDE. Access the documentation ( JavaDocs ) for each of the six APIs mentioned below.

Figures and Tables

Key insights. Please refer to Alag September, 2020 for details.

  • All
  • Figure
  • Table
  • Pipeline

Pipeline

Steps required to generate the results.

Figure 1

COVID-19 clinical trial trends: Longitudinal trends from COVID-19 related clinical trials. The data is plotted across five time points.

Figure 2

Intervention/Drug information. The graph shows the relative frequency of the different intervention/drugs that are referenced in COVID-19 related clinical trials (August 2020).

Figure 3

Intervention Categories. The majority of the interventions used in clinical trials are drugs, other, behavioral, biological, and diagnostic tests

Figure 4

Outcomes.

Figure 5

Phase information

Figure 6

Recruitment status

Figure 7

MeSH information details the most prevalent MeSH terms across COVID-19 related clinical trials.

Figure 8

HPO information portrays the most widely noted HPO terms.

Table 1.

Related drugs, MeSH, and HPO terms using co-occurrences with D018352: Coronavirus infection.

Table 2.

Related drugs, MeSH, and HPO terms using co-occurrences with drug: Hydroxychloroquine.

Data from September 26, 2020

Details of data processed

  • Total number of clincal trials: 352,841
  • COVID-19 related clinical trials: 4,038
  • HPO: nodes: 15,530, Parent-child hierarchy: 19,395, Phenotype to gene: 850,606, Unique genes: 4,366

Java SDK

The JAVA APIs for information about drugs, vaccines, HPO, outcomes, and MeSH terms relevant to COVID-19 clinical trials.

Google Colab Notebook

Reports

Details of COVID 19 related reports across ClinicalTrials.gov:

Item Number
1 Number of unique Interventions/Drugs 4,180
2 Number of unique MeSH terms 691
3 Number of unique HPO terms 263

Data from August 16, 2020

Details of data processed

  • Total number of clincal trials: 348,891
  • COVID-19 related clinical trials: 3,467
  • HPO: nodes: 15,530, Parent-child hierarchy: 19,395, Phenotype to gene: 850,606, Unique genes: 4,366

Java SDK

The JAVA APIs for information about drugs, vaccines, HPO, outcomes, and MeSH terms relevant to COVID-19 clinical trials.

Google Colab Notebook

Reports

Details of COVID 19 related reports across ClinicalTrials.gov:

Item Number
1 Number of unique Interventions/Drugs 3,523
2 Number of unique MeSH terms 622
3 Number of unique HPO terms 254

Data from July 18, 2020

Details of data processed

  • Total number of clincal trials: 345,959
  • COVID-19 related clinical trials: 3,030
  • HPO: nodes: 15,530, Parent-child hierarchy: 19,395, Phenotype to gene: 850,606, Unique genes: 4,366

Reports

Details of COVID 19 related reports across ClinicalTrials.gov:

Item Number
1 Number of unique Interventions/Drugs 3,055
2 Number of unique MeSH terms 572
3 Number of unique HPO terms 229

Data from June 06, 2020

Details of data processed

  • Total number of clincal trials: 341,642
  • COVID-19 related clinical trials: 1,680
  • HPO: nodes: 15,229, Parent-child hierarchy: 18,949, Phenotype to gene: 839,551, Unique genes: 4,315

Reports

Details of COVID 19 related reports across ClinicalTrials.gov:

Item Number
1 Number of unique Interventions/Drugs 1,688
2 Number of unique MeSH terms 323
3 Number of unique HPO terms 123

Data from May 23, 2020

Details of data processed

  • Total number of clincal trials: 340,614
  • COVID-19 related clinical trials: 1,437
  • HPO: nodes: 15,229, Parent-child hierarchy: 18,949, Phenotype to gene: 839,551, Unique genes: 4,315

Reports

Details of COVID 19 related reports across ClinicalTrials.gov:

Item Number
1 Number of unique Interventions/Drugs 1,424
2 Number of unique MeSH terms 293
3 Number of unique HPO terms 122

Data from May 02, 2020

Details of data processed

  • Total number of clincal trials: 332,418
  • Covid-19 related clinical trials: 1,019
  • HPO: nodes: 14,961, Parent-child hierarchy: 18,547, Phenotype to gene: 820,297, Unique genes: 4,312

Reports

Details of COVID 19 related reports across ClinicalTrials.gov:

Item Number
1 Number of unique Interventions/Drugs 1044
2 Number of unique MeSH terms 229
3 Number of unique HPO terms 28