주요 내용 #

Breast cancer와 관련된 Genomic (CNV) Transcriptomic (Gene expression) 유전자변화 종합 분석하고, 그룹화

• Genome-Wide Human SNP Array 6.0 997 tumours, 473 normals
• Illumina HT-12 v3 997 tumours, 144 normals
• Each patients have survival data and clinical features

This data are shared on behalf of METABRIC.

A breast cancer population genomic resource #

Genome variation affects tumor expression architecture #

Expression outliers refine the breast cancer landscape #

Trnas-acting association reveal distinct modules #

Integrative clustering reveals novel subgroups #

Pathway deregulation in the integrative subgroups #

Discussion #

Supplement #

8. Segmentation and copy number alteration calling using CBS #

Normalisation of intensities #

Genome-Wide Human SNP Array 6.0 데이터는 aroma.affymetrix를 써서 normal, tumor 각각 독립적으로 정규화함 (Copy-number estimation using Robust Multichip Analysis (CRMAv2)) - 270 HapMap 데이터 포함.

각각의 샘플에 대해

1. allelic-crosstalk calibration
2. probe sequence effects normalisation
3. probe-level summarisation
4. PCR fragment length normalisation

을 수행하고, log2 intensity 값들을 얻음. 그리고 이어서,

1. genomic position에 따라 probe 정렬
2. replicate probes는 median 값으로 요약
3. missing values는 snapCGH의 loess procedure로 impute
4. QC metrics
   • normalised unscaled standard error
   • relative log expression
   • signal-to-noise ratio of the log2 intensity data

Two pooled references

1. median intensities across the MapMap individuals
2. normal/tumour median intensity values (473 normals)

Calling of copy number alterations and adjustment for tumour cellularity #

DNAcopy 470 normals and 270 HapMap samples

DNAcopy Tomour samples, MergeLevels algorithm

Somatic copy number states:

$$ K_{CNA} = \{HOMD, HETD, NEUT, GAIN, AMP\} $$

Identification of germline CNVs #

$$ K_{CNA} = \{CNVLOSS, CNVGAIN\} $$

Gene-centric alterations #

Probe-level alterations #

Measures of genomic instability #

Identification of recurrent alterations #

9. Segmentation and copy number alteration calling using HMM #

Normalisation of intensities #

6-state HMM for segmentation and discrete copy number prediction in normals #

HMM-Dosage for segmentation and discrete copy number prediction in tumors #

Gene-centric alterations #

10. Comparison of segmentation methods based on MPLA #

Incoming Links #

Related Datasets (Dataset 0) #

• METABRIC

Suggested Pages #

• 0.160 TPX2
• 0.116 eQTL
• 0.114 CPT1A
• 0.033 2015 통계유전학워크숍
• 0.024 RNA editing
• 0.024 GenesWell BCT
• 0.024 IRB
• 0.024 Bioconductor
• 0.024 GWAS
• 0.023 CBS
• More suggestions...