todoDev.txt

? draw labels as divs?
? Put Layout dropdown under the image? (kate)
- multi gene selection, gene "signatures" (czi, holly, olena), ideally sum
- name clusters by top markers (ken harris)
- size dots by expression (ken)
- violin plots (dmitry)
- plot differential expression on x,y, selection versus rest (ken)
- plot most anticorrelated genes. Performance: do z-Scores first, then dot products. anti-correlation better than correlation (ken) + do scatterplot matrix for a few genes (ken)
- identify cells by baby-name (ken)
- get ken harris' clustering to work (ken), it is hierarchical and assigns good names
- can we get the pluritest source code?

x for cell annotations, confusion between display and highlight feature (kate)
x Reset and "sort by" should be buttons, not just text (kate)
x don't show any info if the mouse is not over a cell (kate)
x Add the word "legend" to right hand box (kate)
x work around two finger zoom in old viewer (robert jones)
x make nav tab-active stronger at least (kate)

Ken Harris, phone, May 8 2018:

Ideas:

0) name clusters by top marker gene. assign cluster labels in a hierarchical way
such that Olig2.Pitx2 is a cluster close to Olig2.Six3 

1) combine glyphs and color to show cluster labels. 

2) change size of dot depending on expression level. Do not show expression by color.

3) show differential gene expression on x/y gene plot, not violin plot. Let user select
cells and do differential plot. For each gene, plot 90-percentile, not mean. Plot only
genes that have a high difference in (x-y)/(x+y+1)

4) plot most anti-correlated genes. Correlations can be very fast to calculate, by doing z-Scores
first and then just dot products on the z-Scores. Anti-correlations are more useful
than correlations.

5) Use a scatter plot matrix: a few genes on the x, a few genes on the y, and a scatterplot for each

6) use baby names as cell identifiers. Use official published US baby names list

paper: 
https://www.biorxiv.org/content/biorxiv/early/2018/04/18/143354.full.pdf
code
https://github.com/cortex-lab/Transcriptomics

datasets:
human https://www.ncbi.nlm.nih.gov/pubmed/29727663
brain, SVZ - https://www.ncbi.nlm.nih.gov/pubmed/?cmd=historysearch&querykey=2
immune, HIV infection - https://www.ncbi.nlm.nih.gov/pubmed/29694901
Mouse Brain, Linnarson 600k https://www.biorxiv.org/content/early/2018/04/05/294918 and 
 - data: http://mousebrain.org/dokuwiki/doku.php?id=downloads
Mouse brain, McCarroll, 700k http://dropviz.org/
MCF-7 cell line: https://www.ncbi.nlm.nih.gov/pubmed/29920548

Analysis:
- add https://pluritest.org/ scores