-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.Rmd
52 lines (33 loc) · 3.4 KB
/
index.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
---
title: "Introduction to Slope-Hunter (Under Construction!)"
subtitle: "A robust method for collider bias correction in conditional genome-wide association studies"
author: "Osama Mahmoud"
date: "October 06, 2022"
site: workflowr::wflow_site
output:
workflowr::wflow_html:
bibliography: sh.bib
toc: true
toc_depth: 2
editor_options:
chunk_output_type: console
---
# Welcome to Slope-Hunter website!
Some useful links are as follows:
- [The manuscript published in Nature Communications](https://www.nature.com/articles/s41467-022-28119-9)
- [DOI of the software](https://doi.org/10.5281/zenodo.5617862)
- [R package source](https://github.com/Osmahmoud/SlopeHunter)
- [Website of the corresponding author](http://osmahmoud.com/)
This website layout was inspired by the layout of the [CAUSE website](https://jean997.github.io/cause/index.html).
# Introduction to Slope-Hunter
Studying genetic associations with prognosis (e.g., survival, disability, subsequent disease events) or conditional on a phenotype (e.g., disease incidence) may be affected by selection bias - also termed index event bias or collider bias - whereby selection
on disease status can induce associations between causes of incidence with prognosis.
The [‘Slope-Hunter’ approach](https://www.nature.com/articles/s41467-022-28119-9) is proposed for adjusting genetic associations for this bias. The approach is unbiased even when there is genetic correlation between incidence and prognosis.
Our approach uses advanced machine learning techniques such as unsupervised model-based clustering tailored to theoretical distributions of genetic effects on incidence and prognosis. The [‘Slope-Hunter’ method](https://www.nature.com/articles/s41467-022-28119-9) identifies and utilises the cluster of genetic variants only affecting incidence to estimate an unbiased adjustment factor for collider bias correction even in the presence genetic correlations (e.g., shared genetic pathways as typically observed for many traits including metabolites, cancer risk factors, psychiatric phenotype). The [‘Slope-Hunter’ approach](https://www.nature.com/articles/s41467-022-28119-9) assumes the identified cluster of variants only affecting incidence explains more variation in incidence than any other variant clusters.
# What does this website provide?
- [An easy-to-follow practical tutorial on how to implenment the method](tutorial.html)
- To Appear: Gallery of corrected prognosis GWAS (e.g. disease progression). If you are interested to deposit your prognosis GWAS data after correction in the Slope-Hunter gallery, please do contact [Dr. Osama Mahmoud](http://osmahmoud.com/) on [o.mahmoud@essex.ac.uk](mailto:o.mahmoud@essex.ac.uk?subject="Slope-Hunter: Deposit corrected GWAS data")
# How to cite this work
If you use the [‘Slope-Hunter’ method](https://www.nature.com/articles/s41467-022-28119-9), please do consider to cite both the paper and the software as follows:
- Mahmoud, O., Dudbridge, F., Davey Smith, G. et al. A robust method for collider bias correction in conditional genome-wide association studies. Nature Communications 13, 619 (2022). ([doi: https://doi.org/10.1038/s41467-022-28119-9](https://doi.org/10.1038/s41467-022-28119-9)).
- Mahmoud, O. SlopeHunter: An R package for collider bias correction in conditional GWAS. V1.1.0. (2022). [doi: https://doi.org/10.5281/zenodo.5617862](https://doi.org/10.5281/zenodo.5617862).