Discovery From Data Repositories

21  Download (0)

Full text

(1)

Discovery From Data

Repositories

H Craig Mak  Nature Biotechnology 29, 46–47 (2011)

2013 /06 /10 吳柏誠 陳孝銓

朱唐廷 謝煒騏 李峻霆 徐振庭 葉騰遠

1

(2)

Introduction

Electronic Health Records(EHR)

• Improve care coordination

• Reduce healthcare disparities

• Engage patients and their families

• Improve population and public health

• Ensure adequate privacy and security

11/02/2022

2

(3)

Electronic Health Record Systems

11/02/2022

3

(4)

Health Information Technology for Economic and Clinical Health

Act (HITECH)

• The Health Information Technology for Economic and Clinical H ealth (HITECH) Act seeks to improve American health care deli very and patient care through an unprecedented investment in Health IT (HIT).

Meaningful use The main components of Meaningful Use are:

The use of a certified EHR in a meaningful manner, such as e-prescribing.

The use of certified EHR technology for electronic exchange of health information to improve quality of health care.

The use of certified EHR technology to submit clinical quality and other measures.

11/02/2022

4

(5)

Standardized Administrative Billing Codes

11/02/2022

5

(6)

ICD9

database

11/02/2022

6

(7)

ICD10

11/02/2022

7

(8)

ICD

• International Classification Of Diseases

• WHO(1940)

• Pathogen 、 Pathology 、 Clinical 、 Anatomical location

• Latest version is ICD10

• http://icd9cm.chrisendres.com/index.php

11/02/2022

8

(9)

Genome-wide Association

Study

11/02/2022

9

(10)

Genome-wide association stud y

Basic idea: Screen and test the whole genome for associations with a disease

Use genotyping technologies to assay100,000s of SNPs and relate them to disease

GWAS rely on ”common disease, common variant (of SNP)

Basic idea: A variant (of SNP) has to be common in order to cause a common disease

However, many important disease- causing variants may be rarer than this,and are thus unlikely to be detected with the GWAS approach

GWA studies typically identify common variants with small effect sizes (“common” means >1-5% of population must have the genetic variant )

10

11/02/2022

(11)

GWAS has typically 4 steps

• 1. Selection of large number of individuals with disease, and a comparison group

• 2. DNA genotyping and quality control (QC) of assays and data, check for genotyping errors

• 3. Statistical tests for association between the SNPs (those passing the quality control) and the disease (one test for each SNP, multiple testing…)

• 4. Replication of identified associations in an independent population sample and examination of functionality of identified SNPs

11

11/02/2022

(12)

Limitations to GWAS

• False-positive results (false associations found)

• False-negative results (true associations missed)

• Insensitivity to rare SNP variants

• Requirement for large sample sizes

• Genotyping errors

12

11/02/2022

(13)

What disease associations can we make with a given gene?

 GWAS

Main information sources are genomic databases and electronic medical records (EMR)

PheWAS

( Phenome-wide association scan )

<20 susceptibility genes variants are responsible for >50% of the disease

Potential gene-disease associations could be dicovered

11/02/2022

13

(14)

Introduction

• PheWAS use readily available billing code to replicate sev eral known genotype-phenotype association and suggest s novel possible association.

• Typical EMR system contain diverse data sources, includi ng billing data , laboratory and imaging result, medicatio n records and clinical documentation.

• Structured data V.S. Unstructured data.

11/02/2022

14

(15)

Method

• Selected 5 SNPs which are associated with 7 conditions:

atrial fibrillation, coronary artery disease, carotid artery stenosis, multiple sclerosis, Crohn’s disease, rheumatoid arthritis, and systemic lupus erythematosus.

• Using ICD-9 coding to link genetic data to the phenome.

11/02/2022

15

(16)

Null hypothesis(Ho): 擁有基因型 ’ rs3135388’ 和擁有 ’ MS’,’SLE’

疾病的是不同組人

P-value = 0.05

5.558 (p-value=2.77*10^-6)

11/02/2022

16

(17)

Result

• The PheWAS was able to detect 4 out of the 7 known SN P-disease associations (p<0.05).

• The PheWAS picked up an additional 19 unknown associa tions between the SNPs and diseases

11/02/2022

17

(18)

11/02/2022

18

(19)

Thank You for

Your Attention!!!!!!

11/02/2022

21

Figure

Updating...

References

Related subjects :