Add help popup for SCHEMA gene results

src/browsers/schema/SCHEMAGeneResults.js

@@ -4,6 +4,10 @@ import styled from 'styled-components'
 
 import { BaseTable, TooltipAnchor, TooltipHint } from '@gnomad/ui'
 
+import HelpButton from '../base/HelpButton'
+import StyledContent from '../base/StyledContent'
+import geneResultsDescription from './content/generesults.md'
+
 const Table = styled(BaseTable)`
   min-width: 325px;
 `
@@ -160,7 +164,15 @@ SCHEMAGeneResult.propTypes = {
 
 const SCHEMAGeneResults = ({ results }) => (
   <>
-    <h2>Gene Result</h2>
+    <h2>
+      Gene Result{' '}
+      <HelpButton
+        popupTitle="Gene Result"
+        popupContent={
+          <StyledContent dangerouslySetInnerHTML={{ __html: geneResultsDescription.html }} />
+        }
+      />
+    </h2>
     {results.meta ? <SCHEMAGeneResult result={results.meta} /> : <p>No result for this gene.</p>}
   </>
 )

src/browsers/schema/content/generesults.md

@@ -1,9 +1,7 @@
-# Gene results description [question mark]
-
 This table displays aggregated case-control and _de novo_ counts for the purposes of gene discovery. We first enriched for pathogenic variants by restricting our analysis to ultra-rare protein-coding variants (defined as minor allele count [MAC] ≤ 5) in our data set.  
 
 Given empirically observed exome-wide burden, we focused on protein-truncating or putatively loss-of-function variants (PTVs), defined as stop-gained, frameshift, and essential splice donor or acceptor variants. We additionally analyzed damaging missense variants as prioritized by the MPC pathogenicity score (see [Samocha _et al_. 2017](https://www.biorxiv.org/content/10.1101/148353v1)).   
 
 PTVs and MPC > 3 missense variants (defined as Class I variants) were jointly analyzed in a single case-control burden test. MPC 2 - 3 variants (defined as Class II variants) were analyzed in a separate burden test before meta-analyzed with the Class I burden P-value using Stouffer's weighted Z-score method. Case-control significance was evaluated using a permutation-based Fisher's Exact Test. For genes with case-control P-value < 0.01, _de novo_ Class I and II P values were calculated using the Poisson rate test and meta-analyzed with our case-control test statistic using a Stouffer's weighted Z-score method. The Q-value is the P-value adjusted for the False Discovery Rate.
 
-For a full description and justification of the approach, please consider the main text and supplementary materials and methods of the SCHEMA pre-print manuscript.  
+For a full description and justification of the approach, please consider the main text and supplementary materials and methods of the SCHEMA pre-print manuscript.