20171205

Meeting with Rui:

The fastq files naming: m1 forward, m2 reverse

For FeatureCounts: Count the fragments it wouldn't be much different. unless counting the reads.

• samtools after generated library, 2nd pair shorter ( artificial) samtools view -f 67 It will give us a more precise read counts -q: quality control for alignment mapping

DEseq Fittype you cant choice?

• inform how the experiments are designed with the table?

• build large table before running the DSeq2 ( running the TTseq and RNAseq separately.) First step: read table (format: tabular from Galaxy) from Uppmax -> txt file 12 files -> read them in one step and generate separate count table.

-only for size factors: more differential expression ( internal normalisation because KO normally have few transcription, so if normalise to the same level)

• call the replicates as same sample
• design: compare between samples
• import sizeFactor before calculating
• droplevels ( labeling L) -> running DESeq on only one function: total- KO vs WT
• remove gene counts? different counts between exons vs gene levels

Visualisation of the DE MA- plot: x - mean expression (1st column), y- fold change Volcano plot : x- log2Foldchange, y - P value

Count-based differential exprsssion analysis of RNA sequencing data using R and Bioconductor Anders 2013

IGV view read the mapping