mRNA EDA: PCA and correlation plots of postnormalized expression data, Year123
This document has last been compiled on 2021-12-10 17:36:35.
## Reading in metadata from file: results/Year123/data/root_meta.txt
## Reading in data from file: results/Year123/data/root_normCounts.txt## Total number of root samples: 758
## Total number of genes in root samples: 24939
## No additional bad samples to remove (probably removed during normalization)
## Removing the 72 samples identified as not part of the main experiment, from the root samples
## Variance filter for root samples: Remove genes in the bottom 0.25 quantile of variability
## Total number of genes after limiting to high variable genes : 18704
## After filtering root samples: 686 samples, 18704 genes## Reading in metadata from file: results/Year123/data/leaf_meta.txt
## Reading in data from file: results/Year123/data/leaf_normCounts.txt## Total number of leaf samples: 750
## Total number of genes in leaf samples: 23113
## No additional bad samples to remove (probably removed during normalization)
## Removing the 72 samples identified as not part of the main experiment, from the leaf samples
## Variance filter for leaf samples: Remove genes in the bottom 0.25 quantile of variability
## Total number of genes after limiting to high variable genes : 17334
## After filtering leaf samples: 678 samples, 17334 genesNote that when we are using NORMALIZED data (leaf and root separately), we are using the leaf (root) data file that is the default of bringInDataFunction, the NORMALIZED fpkm values; for this data (leaf and root separately) we are filtering by expression and also on variance within those subsets of samples.
RLE plots
## Day007 Day014 Day021 Day028 Day035 Day042 Day049 Day056
## "#00FF00" "#0BD55F" "#16ABBF" "#2F8EFF" "#6489FF" "#9983FF" "#A66DF2" "#A050DF"
## Day057 Day058 Day059 Day061 Day063 Day064 Day065 Day066
## "black" "black" "black" "black" "black" "black" "black" "black"
## Day068 Day070 Day077 Day084 Day091 Day098 Day105 Day112
## "black" "#D30000" "#A70000" "#991804" "#C46012" "#F0A820" "#FFD01B" "#FFE70D"
## Day119
## "#FFFF00"## Root:##
## Control Postflowering Preflowering
## Day007 5 0 0
## Day014 14 0 10
## Day021 18 0 16
## Day028 18 0 18
## Day035 18 0 18
## Day042 18 0 17
## Day049 14 0 14
## Day056 18 0 14
## Day058 12 0 11
## Day059 10 0 12
## Day061 12 0 12
## Day063 17 18 18
## Day065 12 10 0
## Day066 6 6 0
## Day068 6 6 0
## Day070 18 18 18
## Day077 18 18 18
## Day084 17 14 18
## Day091 16 17 18
## Day098 2 2 2
## Day105 17 18 16
## Day112 1 2 2
## Day119 6 6 6## Leaf:##
## Control Postflowering Preflowering
## Day007 6 0 0
## Day014 14 0 10
## Day021 18 0 14
## Day028 16 0 18
## Day035 18 0 15
## Day042 17 0 18
## Day049 14 0 13
## Day056 18 0 18
## Day058 12 0 12
## Day059 11 0 11
## Day061 10 0 11
## Day063 17 18 18
## Day065 11 8 0
## Day066 6 6 0
## Day068 6 6 0
## Day070 18 18 18
## Day077 18 18 17
## Day084 18 14 18
## Day091 18 18 18
## Day098 2 2 2
## Day105 16 17 16
## Day112 1 1 2
## Day119 6 6 6
PCA of Expression Values
PCA on Leaves
Year123: We first do the PCA on Year123 leaf samples (scaled) colored by different components. The genotypes separate well in the PC1 ~ PC2, while Condition and Time Point (Days) don’t separate well.
Year123: Comment on pair plots of first 4 PCs. While PCs for Condition, Block, Replicate and Time Factor mess up without clustering, Years in PC2 ~ PC4 are a little separated, but not completely.
PCA on Roots
Year123: When the root data is scaled, there appear to be less distinctions between genotypes in the first 2 PCs, though PC2 and PC3 do distinguish them. Years don’t separate in roots in any pairs of the first 4 PCs.
