Lecture 3 Flashcards
What are features of models?
Never exactly like reality
As detailed as necessary
As simple as possible
What types of computational models exist
Mathematical and Statistical Models
What are the 4 assumptions LINE?
xp is a linear function
Errors are Independent
Errors are normally distributed
Errors have equal variance
How does RNA-seq work in steps?
- Fragment RNA
- Sequence Fragments
- Aligne back to genome
- Count reads mapping to gene
How do you prepare a library with Illumina?
Poly-A + RNA capture
RNA fragments primed
cDNA synthesized
3`ends adenlyated
adapter ligation
aplification of fragemnts
What is the Phred Score?
P = 10 ^(-Q/10) P = Propability base call incorrect Q = -10log(10)P
What is read trimming?
Chopping of ends of Read Quality Graphs
What is Read mapping?
Alignment of read to base gene sequence
What determines read count per base?
Expression level
Sequencing depth
Mapability
Noise
What determines read count per transcript?
Expression level
Sequencing depth
Mapability
Noise
Transcript length
Splicing
What are two methods for RNA enrichment?
poly A enrichment
purify transcripts with poly-A tail
enrich for mRNA
deplete non coding RNA
ribo-minus
remove rRNA
keep non coding tho
What is FPKM and how do you calculate it?
FPKM = n(i) / (l(i) * N) // n(i) = number of fragments for transcript i
l(i) = length of transcript i
N = 1 / 1000000 * Summe n(i)
What is TPM and how do you calculate it?
n(i2) = n(i) / l(i)
S = 1 / 1000000 * Summe n(i2)
TPM = n(i) / S
Whats the difference between TPM and FPKM
TPM is the same for each sample while FPKM may vary since it depends on the length of the fragment
How to determine differentially expressed genes?
linear model:
π = π + π½1 * π₯π‘reat +π½2 * π₯πatch +π½3 * π₯π ex
Y = read counts in a given sample
Β΅ = average read counts
xtreat = treatment condition (e.g. knock-out or drug treatment)
xbatch = batch membership (account for batch effects)
xsex = sex of the animal (account for sex effects)
