5.1 ChIP-SEQ

Question 1

What are some applications of sequencing?

Answer 1

• WGS variant calling
• WGS metagenome
• RNAseq Transcriptomes
• Chipseq protein DNA interactions

Question 2

What are histone modifications?

Answer 2

a type of epigenetic modification that acts to reinforce open or closed chromatin conformations

Question 3

What do enhancers do?

Answer 3

regulate gene activity; are defined and controlled by epigenetic state

distal and proximal

Question 4

What are ways genomes can be sheared?

Answer 4

1. sonication
2. enzymes
3. transposons

Question 5

What is ChIPseq?

Answer 5

a sequencing based approach used to measure histone modification patterns in the genomes

Question 6

What are each of these marks characteristic of:

1. H3K4me3
2. H3K4me1
3. H3K27ac
4. H3K36me3
5. H3K9me3
6. H3K27me3

Answer 6

1. active promotors
2. enhancers
3. active enhancers
4. elongation transcription
5. repressive/heterochr
   6 . repressive

Question 7

What are the 4 main steps in the ChIp-seq analysis workflow

Answer 7

1. 2 libraries of generated: (1) library made using ChIPseq & abs against histones ( 2 ) control library called the input ; fastq
2. Align and pair sequence reads;
   tool: BWA mem
   file: .SAM
3. Convert to binary format, sort
   tool: SAMMAMMA or SAMtools
   file: .sorted.bam
4. Generate density tracks and call peaks
   tool: MACS2
   file: .bedgraph, .bed

Question 8

What are BigWigs?

Answer 8

binary, indexed version of BEDgraph files that are loaded directly into the browser for visualization

AKA compressed binary version of BED and BEDgraph files

Question 9

What file format is used to visualize Chip data?

Answer 9

Bed and BedGraph files

Question 10

What are the 3 required fields in a BED file?

Answer 10

1. chrom: name of chromosome
2. chronStart: start position of the features in the chromosome or scaffold, 0 -indexed
3. chromEnd: ending position of the features in the chromo or scaff

Question 11

How are the coordinates in BED file reported?

Answer 11

Half open; closed start and open end [A, B)

Question 12

What is a BEDgraph file?

Answer 12

A 4 column BED file with a track header; It contains a track definition line to set browser parameters

The columns include the same 3 as BED file + a 4th dataValue column

Encode quantitative data like signal amplitude

Question 13

What are Key considerations in ChIP-seq

Answer 13

1. antibody specificity and sensitivity
2. Which marks to profile
3. required sequencing depth

Question 14

What is a wiggle format?

Answer 14

Way of encoding single nucleotide information; single based resolution bed format

Question 15

Why might the required sequencing depth differ in ChIpseq?

Answer 15

• feature prominence; is something is prominent, need more sequencing reads to cover those events (AKA mark occupany)
• 50M read pairs for punctate marks
• 100M read pairs for broad marks

Question 16

What is the input/control used in ChIpseq?

Answer 16

Input library consists of the sheared DNA prior to IP; the control is used for background correction

Question 17

What is iCHIP and what are the advantages of it?

Answer 17

Indexed are added at the ligation step allowing for multiple experiments to be pooled together before IP
Advantage: can use smaller # of cells which might be rare

Question 18

How is the quality of a CHIPseq sequencing run file evaulated?

Answer 18

1. sequence quality
2. Library quality
3. IP quality

Question 19

How is sequence quality deterined?

Answer 19

By FASTQC

Question 20

How is library quality determined?

Answer 20

By the diversity of IP’d fragments in library used as measure of library quality; the more diverse fragments present, the more represented of the target

Question 21

What is the relationship between library diversity and PCR duplicate rate

Answer 21

inversely correlated

Question 22

What are PCR dups and how are they created?

Answer 22

Are reads that have identical start and stop positions relative to a reference

NOT created by clonal amplification but caused by PCR cycles used to amplify library after adapter ligation

Question 23

What do the presence of PCR duplicates mean?

Answer 23

That the library generated is less diverse and therefore lower quality

Question 24

Why might duplicates be bad?

Answer 24

If initial starting material is low, duplicates can lead to overamplification of the material before sequencing. Any biases in PCR will compound this problem and lead to artificial peak calls

In Chip we are trying to look for event frequencies

Question 25

What are good duplicates?

Answer 25

biological duplicates are expected because small parts of the genome that are enriched by IP are being sequenced.

Question 26

What happens when u remove biological duplicates?

Answer 26

The dynamic range of the CHIP-seq signal for that region is compressed.

Question 27

How can biological duplicates be identified?

Answer 27

1. Add 3 random N's to adaptor, where rest of the adaptor is a standard adaptor sequence 2. These 3 sequences are unique to each fragment before PCR 3. Use these unique identifiers to tell difference between biological and PCR duplications

Question 28

How does the appearance of biological duplicates compare to PCR duplicates?

Answer 28

Biological dups will have multiple overlapping reads with offsets PCRdup will look like reads stacks that align perfectly

Question 29

What are two common measures of IP quality?

Answer 29

FRIP: Fraction of sequencing aligning within called peaks Domain reads

Question 30

What are domain reads?

Answer 30

- fraction of sequencing alignments within a defined set of genomic regions - ex: Gene, promoters, - Specific to the IP'd target

Question 31

Why do you want to look at the number of reads that align to a peak?

Answer 31

Tells you what the signal to background is

Question 32

What is the purpose of MACS2?

Answer 32

identify regions of enrichment comparing a treatment (IP) to a control (input)

Question 33

What are the firs 4 general steps in MACS2

Answer 33

1. Taking BAM file, scan along chromosome to find significantly enriched bins (2x average fragment size) with counts/mfold higher than random genome average 2. For 1000 randomly choose enriched bins calculate difference between distribution max of reads on + and - strand = d 3. shift add + reads by +d/2 and (-) reads by -d/2 4. Scale control experiment to the same number of reads as chip seq

Question 34

What are regions that are high enriched in ChipSEQ?

Answer 34

Usually artifacts that are PCR duplicates

Question 35

Why is d calculated and reads shifted by d/2 in chipseq?

Answer 35

The nucleosome inhibits DNA sharing at the true peak

Question 36

What kind of distribution does MACS2 use for read count?

Answer 36

Poisson distribution

Question 37

In MACS2 what does lambda represent?

Answer 37

The max mean number of reads that align with control bin (the expected number of reads in that window); calculated from input Lamda = max value obtained from this control

Question 38

How is a p-value calculated in MACS2?

Answer 38

p-value calculated using IP-mean and poisson distribution for lamda

Question 39

What is the False Discovery Rate?

Answer 39

The expected proportion of discovered that are falsely rejected

Question 40

What are the inputs for MACS?

Answer 40

1. treatment file (required); BAM or BED | 2. control/input

Question 41

How can CHIPseq data be visualized?

Answer 41

bedgraphs can be visualized using IGV