Friday August 1 week's notes

2014-07-29

ApGxL

Completed DGE protocol: Post-SPIA Modification I and II, purification using AMPure beads. Did this in the molecular lab to prevent cDNA contamination of future DGE samples.

MtGiraffe

Meeting with JH.

VCF files - reporting positions with single nucleotide and missing data (“./.”) as SNPs

filter SNP file to true SNPs
another interesting genes: RBOHC Medtr3g098350.1
next to two other similar genes
mis-expressed at seedling stage and in senescing nodule
lower expression, longer roots

2014-07-30

ApTranscriptome

Incorporating comments from co-authors

RStudio webinar

Grammars and graphics of data science
Hadley Wickham
collect [tidy] -> analyse [compose] -> communicate
focus on analysis
tidy data
analysis
transform
visualise
model
iterate!
visualization does not scale - summarise with models
however, models have assumptions so cannot reveal unexpected patterns
tidy data: tidyr
tidy data easy to work with
store variables in consistent way, always as columns
transform: dplyr
bottlenecks in data analysis
cognitive -> computational
thinking: describe data manipulation problem
5 functions
- filter: keep rows matching criteria
- select: pick columns by name
- arrange: reorder rows
- mutate: add new variables
- summarise: summarise variables
Demo: nycflights13
group data using group_by
pipelines
- pipe operator: %>%
remote sources; connect to databases
- will generate SQL code automatically!
- avoid switching between SQL and R code because small but annoying differences
- browseVignettes(package = "dplyr")
ggVis for visualizing data in R (Winston Chang)
interactive data visualization
communicate: explain, deploy, display
allows for input sliders; binwidth, opacity
ggVis in Shiny for reactivity
- tooltip that gives info on points!
- fill.brush to select points, highlight in other charts

Code from presentation on github

ApGxL

Completed DGE protocol: Post-SPIA Modification I and II, purification using AMPure beads. Did this in the molecular lab to prevent cDNA contamination of future DGE samples.

Qubit results:

Samples	ng/ul
01-B-HS	9.03
02-A-HS	9.22
02-C-HS	10.7
04-A-HS	>60

Update 2014-08-06: re-ran 04-A-HS at 10x dilution (0.2ul in 1.8ul H₂O).

Samples	ng/ul	stock (ng/ul)
04-A-HS	5.83	58

2014-07-31

ApTranscriptome

Edits to manuscript

ApGxL

Qubit RNA assay for sample extracted with RNAzol

Sample	ng/ul
04-A-CO	<20
O5-B-CO	37.6
10-A-CO	<20
1O-B-CO	45.0
12-A-CO	<20
22-A-CO	<20
22-B-CO	4.4
22-C-CO	<20

RNA in only 3 of 8 samples…disappointing.

Also checked RNA concentration of the DGE preps

Sample	ng/ul
02-C-HS	2.4
04-A-HS	9.9

RNA contamination in cDNA from DGE samples…is this a problem?

2014-08-01

ApGxL

Emailed Ashesh at Nugen to ask about DGE protocol.

How much RNA was used as input into the system? How was the RNA isolated? What is the RNA quality (RIN, 260/280 and 260/230 ratio)? How was the amplified cDNA purified and what was the elution volume? The reason that I am asking is that we would expect ug quantities of cDNA. Your concentration of > 9 ng/ul seems to be low. Did you do bioanalyzer analysis on the amplified cDNA? Can you send me the traces?

It is good to hear from you. I have a few questions about the cDNA amplification. How much RNA was used as input into the system? How was the RNA isolated? What is the RNA quality (RIN, 260/280 and 260/230 ratio)? How was the amplified cDNA purified and what was the elution volume? The reason that I am asking is that we would expect ug quantities of cDNA. Your concentration of > 9 ng/ul seems to be low. Did you do bioanalyzer analysis on the amplified cDNA? Can you send me the traces?

To answer your questions; while this system uses poly(A) priming you may see some rRNA read. This is because rRNA sequences can have poly(A) stretches that may be primed and amplified. This is also organism dependent since we see higher rRNA when using C. elegans total RNA compared to human total RNA. You will see these sequences in your library but it should be a small percent of the total number of reads. Secondly, library construction requires the presence of double-stranded DNA so the presence of RNA should not interfere with the library construction.

Are you doing paired-end or single end reads? For sequencing 100bp reads, I would recommending fragmenting the cDNA to 200 bp. The cDNA should be purified after fragmentation. When using the Rapid library system, I would recommend using 1ug of the fragmented cDNA as input into the system. Additionally, I would like to remind you that a qPCR library quantification is required to accurately quantify the libraries produced with the Rapid system.

So…the SPIA amplification doesn’t seem to be working to expectations…

ApTranscriptome paper