boat ball day

Question 1

Summarise how DNA is extracted – how are proteins removed? How is DNA purified with the column? How is it eluted?

Answer 1

Lyse cells, Proteinase K and denaturation at 65C, centrifuge to separate protein (pellet) from supernatant (DNA), and precipitate out. Then salt so phosphate forms H bonds with silica for purification and elute with lower salt conc.

Question 2

How does the Nanodrop measure purity?

Answer 2

Measures UV absorbance.

Question 3

What does 260/280 ratio measure? What do you expect for
1) pure DNA
2) pure RNA
3) contamination

Answer 3

dna/rna (260) protein contamination (280 ab)

1.8
2.0
>1.8

Question 4

What does 260/230 measure?
What is expected and contaminants?

Answer 4

230 = contamination
2-2.2 = pure DNA/RNA
>2.2 = contamination

Question 5

How does the PicoGreen assay quantify DNA and how is the DNA quantification measured? How is it specific?

Answer 5

Fluorescent probe binds minor groove dsDNA - standard curve to quantify concentration.

Question 6

What does an adaptor do and how is it added to DNA?

Answer 6

ligated - barcode (sampleid) and primers

Question 7

What are the 3 steps of PCR? What else do you need in PCR?

Answer 7

1. denature 95
2. anneal 55
3. extend 72

dNTPs, polymerase,

Question 8

Why do we need to amplify DNA?

Answer 8

Enough DNA to detect signal in qPCR - so less noise

Question 9

What does qPCR tell you? What does the graph look like? How is it quantitative? How do you find the concentration of your DNA sample – what is needed in the experimental procedure?

Answer 9

concentration of dna in sample. signal vs cycles. need a standard curve from dilutions of known concentrations and addition of fluorescent probe.

Question 10

Why do we amplify the 16S rRNA gene? How are the primers designed?

Answer 10

gene in all bacteria - can be mapped to database to identify taxonomy. variable regions flanked by highly conserved - use as primers.

Question 11

What is the amplicon length of V4-V5 region?
What is the read length of Illumina platforms? Why is paired-ended reads needed?

Answer 11

450

250

pair end can do both ends and then marge in vsearch pipeline - high quality, more accurate

Question 12

What does the FASTQ file contain?

Answer 12

sample id & seq

Question 13

What is an OTU and what does it group? What is the difference between OTU and zOTU?

Answer 13

operational taxonomy unit - groups reads together with similar seuqneces in an attempt to cluster same species of bacteria

otu = 97% sequence similarity and zOTU = 100%

Question 14

What is VSEARCH?

Answer 14

open search tool used in pipelines

Question 15

What is merging and why is it necessary? What is the minimum overlap required? Why do we trim?

Answer 15

need to merge paired ends into 1 sequence - 25-30bp - need to trim as 3’s are noisy, and can increase accuracy to only include high accuracy merged region which can still be used for taxonomic recognition more than inaccurate

Question 16

How does a MaxEE of 1.0 filter the reads?

Answer 16

max estimated error = 1 - all reads calculate estimated error and if >1 then removed. sum of probabilities that base is wrong so EE 1.0 = amongst 100bp only 1 is incorrect.

Question 17

What is dereplication? Why is this important?

Answer 17

collapses identical sequences into just 1 - to only keep just 1 = more efficient and smaller files/tables to work with

Question 18

What is SWARM?

Answer 18

algorithm that clusters similar sequences into OTUs (97% similarity) iteratively

Question 19

What is UNOISE? Why is this better than SWARM?

Answer 19

Denoising algorithm that infers zOTUs (100% sequence similarity) or ASVs using error modelling and denoising with high resolution 100%match

Question 20

What is a singleton and why are they removed?

Answer 20

Singleton = sequence that only appears once - more likely error of PCR/contamination rather than true sequence

Question 21

What is a chimera and why are they removed?

Answer 21

PCR artefact where dna strands incompletely synthesised act as primers and make hybrid dna sequences - error

Question 22

What does the final OTU table look like?

Answer 22

OTU on the rows, and samples on the columns e.g. sample 1 has zotu1 of 2 = 2 sequences detected belonging to zotu1

Question 23

What is a metadata file made up of?

Answer 23

sample id with group variables e.g. sample 1 = location ucl

Question 24

How is taxonomy assigned to the otu table? what does this?

Answer 24

sequence associated with the zotu mapped against database qiime2

Question 25

What is qiime2 used for?

Answer 25

open source tool for microbiome analysis - including alpha and beta diversity

Question 26

What is a rarefaction curve? how do you interpret it?

Answer 26

plot showing x = sequencing depth bp and y = OTU number plateau = all diversity captured in the selected seq depth no palteau = more diverse

Question 27

before sequence analysis, why is the feature/otu table normalised?

Answer 27

varying library sizes - normalise to avoid skewing data to favour larger libraries as more rich - so use lowest number of reads in a sample

Question 28

what does alpha diversity measure?

Answer 28

measures the diversity within a group with respect to richness (no. tax groups) and evenness (distribution of abundances of tax groups)

Question 29

what does beta diversity measure?

Answer 29

measures diversity between the samples in a group using distance metrics - how close they are in space looking at the variation between samples

Question 30

what does Shannon measure?

Answer 30

both richness and evenness - high shannon = many taxa with even abundance - every species in richness including low abundance

Question 31

what does simpson measure?

Answer 31

evenness - high simpson = high dominance of few taxa - doesnt consider rare taxa

Question 32

what does faith pd measure?

Answer 32

phylogenetic RICHNESS - branch length (smaller = more closely related) between all taxa in tree - demonstrates how bacterial microbiomes differ in how related they are/evolutionary relationships rather than just counting abundance of taxa - can look at pressures of evolution

Question 33

what does chao1 measure?

Answer 33

richness - uses singletons (which are removed in our pipeline) to estimate number of rare taxa - more sensitive than shannon

Question 34

what is the kruskal wallis test used for? what does p < 0.05 mean what are teh assumptiosn?

Answer 34

compare 3+ groups to assess whether they come from the same distribution (null assumption) significant differences amongst at least 1 group not due to random chance continous, not normally distributed data (even distribution and variance is same)

Question 35

what does weighted unifrac distance measure?

Answer 35

accounts for abundance to detect dominant species and weights evolutionary branches on abundance

Question 36

what does unweighted unifrac distance measure?

Answer 36

looks at which taxa are present/absent in samples - sensitive to rare taxa

Question 37

what does bray curtis dissimilarity measure?

Answer 37

dissimilarity between 2 samples on taxa presence and abundance where 1 = different and 0 = same

Question 38

what does jaccard measure?

Answer 38

similarity between two samples on taxa present (not abundance) where 1 = same and 0 = different

Question 39

why are ordination methods used?

Answer 39

difficult to visualise distance metrics - ordination reduces the resolution so can fit into 2d plot for visualisation to identify patterns

Question 40

how does PCOA work? what does PC1 and PC2 represent? if PC1 = 10% what does this mean? whats the difference between pcoa and pca?

Answer 40

transforms a distance matrix between samples into set of axis called principle components PC1 = the largest component responsible for variation PC2 = the second largest component responsible for variation then PC1 explains 10% of the variation pcoa = non euclidean and pca = euclidean

Question 41

what does permanova measure and what is it used for? what is the assumption? what is the null and alternative hypothesis?

Answer 41

evaluates whether the composition of a microbiome group differs in multivariate space e.g. in a PCoA e.g. does the microbiome differ between ucl vs home? not normally distributed - true to microbiome data. , assumes equal variance between groups null = centroids/groups all even in space alt = at least one group significantly differs - group effect not explained by random variance

Question 42

what does denoising include? why do we denoise?

Answer 42

after quality filtering - derep, chimera/singleton removal and assign zOTUs remove errors, remove replications for efficiency, remove chimeras (errors seq), and identify otu with sequence species

Question 43

what does a dispersion test measure? why does this need to be done? what does p < 0.05 mean? what about p > 0.05?

Answer 43

checks whether the variability between groups is equal permanova assumes that dispersion between groups are equal (null) and if they arent then it's due to a group effect - but doesnt account for variability so dispersion tests measures whether the difference seen is due to the group or due to variability within the group groups have equal dispersion so okay to use permanova groups have different dispersions & variability - so if performed permanova it could bias the result to saying group difference when just unequal variance

Question 44

how would you interpret low dispersion on a pcoa plot? higher dispersion?

Answer 44

tight cluster = samples are similar more variability between samples

Question 45

how do you interpret if two groups occupy similar spaces on pcoa axes?

Answer 45

microbial communities are similar based on the beta diversity metric used - e.g. unifrac would be phylogenetic similarities

Question 46

how would you plot alpha diversity? why are error bars important?

Answer 46

box plots shows the variation within the group - more variation = less confidence and if error bars overlap then no significant differences

Question 47

how could you better look into taxa bar plots what do you need to consider with taxa bar plots? what is abundance testing?

Answer 47

make box plots with distribution and variance for each species across all groups number of species - is it too messy, figure legend - is it relative abundance? error bars? colours? interpretation hard? LEfSe - tests how abundant a taxa is in one group than another

Question 48

what to look for when analysing beta diversity plots?

Answer 48

clustering, outliers, does PC1/PC2 explain for variance between different groups? what types of distance matrix used and what does this mean?

Question 49

what to look for when analysing alpha diversity plots? what happens if the rarefaction curve seq depth is too small? too large?

Answer 49

error bars, clear legend/groups, is it median shift and IQR? how are the data poitns dispersed, does the rarefaction curve capture enough diversity? what is the measure? doesnt accurately capture the depth of diversity at sequencing level lot slower for qiime2, can skew abundance such lower rare taxa not detected

Question 50

what do the following SWARM settings mean in clustering: -d -f -z why is unoise/usearch better?

Answer 50

distance threshold = number of nucleotide differences that can differ between sequences - default is 1 fastidiuous - more strict algorithm - sequences have high similarity outputs sequences of each swarm/OTU cluster faster, higher resolution (1bp level), exact error corrected zOTUs

Question 51

what are the 3 types of alpha diversity?

Answer 51

qualitative - presence/absence quant - abundance weighted phylo - evo relationships

Question 52

how does the bradford assay work?

Answer 52

coomassie brilliant blue binds aromatic res in proteins (phenylalanine, tryptophan) - need dilution standard curve to measure protein conc - spectrophotometry measure blue absorbance from brown

Question 53

what is the peptide fluorometric assay?

Answer 53

flouorometric reagent reacts w primary amines -> fluorescence propoertional to peptide content

Question 54

summarise how rna is extracted.

Answer 54

use beads to smash open cell membranes, denature rnases and proteins, silica column binds rna and eluted and treated with dnase to remove dna

Question 55

how much dna/rna would you expect (ng/ul) from: hospital swap fecal bacteria tissue

Answer 55

1-20 10-50 20-100 50-200