1 of 40

Statistical Genomics & Genetics

Johns Hopkins Biostatistics

February 21, 2020

Stephanie Hicks

Assistant Professor, Biostatistics Department�Faculty Member, Johns Hopkins Data Science Lab��stephaniehicks.com�Twitter: @stephaniehicks

2 of 40

what makes us diverse?

slide adapted from alyssa frazee

3 of 40

how does this happen?

slide adapted from rafa irizarry

4 of 40

how does a healthy cell become a cancer cell?

5 of 40

AUCAGUCGAUCACCGAU

transcription

RNA

translation

protein

ACTGACCTAGATCAGTCGATCGATCGTATACGATTACAAAATCATCGGCAT

DNA

central dogma

slide adapted from alyssa frazee

6 of 40

ACTGACCTAGATCAGTCGATCGATCGTATACGATTACAAAATCATCGGCAT

DNA

genetics

phenotype

7 of 40

Different genomes, different phenotypes

Sloth

Human

8 of 40

ACTGACCTAGATCAGTCGATCGATCGTATACGATTACAAAATCATCGGCAT

DNA

genetics

9 of 40

AUCAGUCGAUCACCGAU

transcription

RNA

translation

protein

ACTGACCTAGATCAGTCGATCGATCGTATACGATTACAAAATCATCGGCAT

DNA

central dogma

slide adapted from alyssa frazee

10 of 40

AUCAGUCGAUCACCGAU

transcription

RNA

translation

protein

ACTGACCTAGATCAGTCGATCGATCGTATACGATTACAAAATCATCGGCAT

DNA

genomics

M

M

M

slide adapted from alyssa frazee

  • phenotype

11 of 40

Taub, Rucinski, Chatterjee, Zhao

�Hansen, Hicks��Ji, Hansen��Hicks, Ji, Hansen, Leek��Ruczinski

DNA-seq�

DNAm�

ChIP-seq�

RNA-seq�

Protein

Genome

Function

12 of 40

Slide courtesy: Ben Langmead

13 of 40

ACTGACCTAGATCAGTCGATCGATCGTATACGATTACAAAATCATCGGCAT

DNA

data generation

14 of 40

data generation

GATCGATCGTATACGAT

Fragments

ACTGACCTAGATCAGTC

TACAAAATCATCGGCAT

ACTGACCTAGATCAGTCGATCGATCGTATACGATTACAAAATCATCGGCAT

DNA

15 of 40

data generation

GATCGATCGTATACGAT

Fragments

ACTGACCTAGATCAGTC

TACAAAATCATCGGCAT

Reads

TACAAAATCA

AGATCAGTC

GATCGATCG

ACTGACCTAGATCAGTCGATCGATCGTATACGATTACAAAATCATCGGCAT

DNA

16 of 40

@22:16362385-16362561W:ENST00000440999:2:177:-40:244:S/2

CCAGCCCACCTGAGGCTTCTTTTTCCTTCCCAAGCCACATCACCATCCTGGTGGAACTCTCCTGTGAGGACAGCCA

+

GGFF<BB=>GBGIIIIIIIIIIIIIIEGEHGHHIIIIIIIIHFHBB2/:=??EGGGEGFHHIHHEDBD?@@DDHHD

@22:16362385-16362561W:ENST00000440999:3:177:-56:294:S/2

GCGTGAGCCACAGGGCCCAGCCCACCTGAGGCTTCTTTTTCCTTCCCAAGCCACATCACCATCCTGGTGGAACTCT

+

@=ABBBBIIIIIIIIHHGGGGIIDBDIIIIIIGIIIIHIIIIHFDD@BBDBGGFIDEE8DCC/29>BGFCGHHHGF

@22:16362385-16362561W:ENST00000440999:4:177:137:254:S/1

TCACCATCCTGGTGGAACTCTCCTGTGAGGACAGCCAAGGCCTGAACTACCTGCaGTGGGGAGCACCTCAGGGTTT

+

DDGBBCGGGIGGGBDDDHIIGGDGD77=BDIIIIIIIIFHHHHIIIHEFFHGGDD8A>DEGHHIFDDHH8@BEDDI

@22:16362385-16362561W:ENST00000440999:5:177:68:251:S/2

AGGGTTTGCCCAGGCAACCAGCCAGCCCTGGTCCAAGGCATCCTGGAGCGAGTTGTGGATGGCAAAAAGACNCGCC

+

HIGHIHFHEGE4111:.;8@?@HDIIIIIIIEGGIHHHIIGA?=:FIIIDD8.02506A8=AC#############

@22:16362385-16362561W:ENST00000440999:6:177:348:453:S/1

AAGGCCTGAACTACCTGCGGTGGGGAGCACCTCAGGGTTTGCCCAGGCAACCAGCCAGCCCTGGTCCAAGGCATCC

+

B9?@8=42:E@GDEDIIIIIGGHIIIFBEEAGIIDIIDHHGGHIIEGEIIIIIHIHFHFFEEFGGGGGB88>:DGH

@22:51205934-51222090C:ENST00000464740:132:612:223:359:S/2

GGAAGTATGATGCTGATGACAACGTGAAGATCATCTGCCTGGGAGACAGCGCAGTGGGCAAATCCAAACTCATGGA

+

IIEHHHHHIIIIIIIHGGDGHHEDDG8=;?==19;<<>D@@GGGIIHIIHGGDDHGBA=ABEG@@DFCCAA<:=>8

@22:51205934-51222090C:ENST00000464740:125:612:-1:185:S/1

TGGAGTGCGCTGCGGCGCGAGCTGGGCCGGCGGGCGTGGTTCGAGAGCGCGCAGAGTCCAGACTGGCGGCAGGGCC

+

GGFF<BB=>GBGIIIIIIIIIIIIIIEGEHGHHIIIIIIIIHFHBB2/:=??EGGGEGFHHIHHEDBD?@@DDHHD

@22:16362385-16362561W:ENST00000440999:3:177:-56:294:S/2

GCGTGAGCCACAGGGCCCAGCCCACCTGAGGCTTCTTTTTCCTTCCCAAGCCACATCACCATCCTGGTGGAACTCT

+

GGFF<BB=>GBGIIIIIIIIIIIIIIEGEHGHHIIIIIIIIHFHBB2/:=??EGGGEGFHHIHHEDBD?@@DDHHD

@22:16362385-16362561W:ENST00000440999:3:177:-56:294:S/2

GCGTGAGCCACAGGGCCCAGCCCACCTGAGGCTTCTTTTTCCTTCCCAAGCCACATCACCATCCTGGTGGAACTCT

+

​

​

​

​

billions more

17 of 40

N =

SAMPLE SIZE

18 of 40

N =

($ YOU HAVE)

($ PER SAMPLE)

19 of 40

$ per (human) Genome

http://www.genome.gov/sequencingcosts/

20 of 40

21 of 40

All the data

22 of 40

ACTGACCTAGATCAGTCGATCGATCGTATACGATTACAAAATCATCGGCAT

DNA

genetics

phenotype

23 of 40

Rare and common variants → relative risk of disease

24 of 40

Ingo Ruczinski

Family study (rare variants)�Goal: Identify highly penetrant disease variants

by sequencing distant relatives

25 of 40

Genome-wide association study (common variants) �Telomere length from 75,000 individuals

Manhattan plot showing peak genetic signals

Margaret Taub

26 of 40

TACAAAATCA

AGATCAGTC

GATCGATCG

All the dataz

+

what we do

27 of 40

TACAAAATCA

AGATCAGTC

GATCGATCG

All the dataz

+

experimental design

experimental design

28 of 40

TACAAAATCA

AGATCAGTC

GATCGATCG

All the dataz

+

experimental design

preprocessing

+

normalization

29 of 40

TACAAAATCA

AGATCAGTC

GATCGATCG

All the dataz

+

genomic �data science

30 of 40

Ni Zhao

Measuring the impact of the microbiome�MiRKAT: kernel methods for associating microbiome data with phenotypes of interest

31 of 40

Kasper Hansen

De-noising DNA methylation data

32 of 40

Kasper Hansen

De-noising DNA methylation data

33 of 40

Kasper Hansen

Understanding changes in DNA methylation�in colon cancer

34 of 40

“...10 billion observations (or cells) by 2020”

35 of 40

Stephanie Hicks

Modeling single-cell RNA-sequencing data

(single-cell) RNA-seq data are nonnegative integers

36 of 40

Stephanie Hicks

Generalized Principal Components �Analysis (GLM PCA)

PCA

GLM PCA

37 of 40

In a nutshell

Interesting, intellectually challenging, �scientifically important problems��Big and complex data��Unique contributions to both science and statistics

slide adapted from kasper hansen

38 of 40

Two major roles for statisticians

As safeguards against mistakes

​

As engines of discovery

slide adapted from hongkai ji

39 of 40

Outside department: �JHU genomics (broader hopkins community) �meets 2x month��Inside department: �Lots of working groups in biostats for �both statistical genetics and genomics

Group meetings

40 of 40

Taub

Leek

Ji

Hansen

Hicks

Zhao

Ruczinkski

Chatterjee