Biomarkers Associated With Therapeutic Outcomes (BAWTO) In Prostate Cancer

github.com/insilica/bawto

Scripts and gists for creating/analyzing data synthesized from prostate cancer clinical trial publications.

Three sysrev projects were created for this analysis:

  1. sysrev.com/p/63101: Screening clinicaltrials.gov
  2. sysrev.com/p/68027: Screening pubmed abstracts
  3. sysrev.com/p/70431: Cancer trial data synthesis

Fig1 - Prisma diagram

bawto/figure-scripts/fig1-prisma.html
Notebook takes you step by step through generation of a prisma diagram in R.

Fig2 - Timeline

bawto/figure-scripts/fig2-timeline.html
Building a timeline for prostate cancer biomarker clinical trials with vistime.

Fig3 - Trial Inclusion Biomarkers

bawto/figure-scripts/fig3-heatmap.html
Generating a heatmap to evaluate biomarker usage in cancer clinical trials.

Fig4 - Hazard ratios in a Treemap

bawto/figure-scripts/fig4-HR-treemap.html
Use plotly to generate a treemap for intervention trial hazard ratios. Learn which biomarkers contribute to the observed hazard ratios in a cohort.

Fig5 - PSA response forest plot

bawto/figure-scripts/fig5-psa-forest.html
The Meta package can analyze proportions of patients that have a PSA progression within a given timeframe. Learn to create forest plots with meta::forest.

Fig6 - Adverse Events

bawto/figure-scripts/fig6-adverse-event.html
Clinicaltrials.gov reports some trial adverse events, but others can only be extracted from the published literature. Use the patchwork package to build a grid of figures describing marginal distributions of adverse events as a function of trial and therapy.

Fig7 - Model based trial partitions

bawto/figure-scripts/fig7-models.html
Clinicaltrials.gov reports some trial adverse events, but others can only be extracted from the published literature. Use the patchwork package to build a grid of figures describing marginal distributions of adverse events as a function of trial and therapy.

Explore published trial data with SQL

bawto/figure-scripts/statements.html
Write a few basic queries to understand which trials had the greatest or lowest adverse event rates, or which therapies had the longest survival time. go to notebook