TraVA is a database of gene expression profiles in Arabidopsis thaliana developmental stages, organs and parts based on RNA-seq analysis. Profiles of 25 706 protein-coding genes (gene models are according to TAIR10 genome release) in 79 samples (complete list of samples with their detailed description can be found here) are presented. There are two main blocks of analysis: 1) gene expression levels 2) differential gene expression For any analysis, gene identifier should be entered in the window "Enter gene id". Acceptable identifiers are TAIR identifiers (ATXGXXXXX) or gene names. In latter case, use only abbreviated names (for example, LFY but not LEAFY, SAP but not STERILE APETALA). If gene is known under different names, only TAIR gene name, not an alias, is acceptable (for example, AGL8 for AGAMOUS-LIKE8, known also as FRUITFULL, but CAL for CAULIFLOWER, known also as AGAMOUS-LIKE10). For more information on gene names refer to https://www.arabidopsis.org/portals/nomenclature/guidelines.jsp

To start enter your gene ID in ATxGxxxxx format.

Working with gene expression levels.

This block allows you to see the expression levels of a given gene or multiple genes in each sample, without making comparisons between samples. In case of multiple genes, though theoretically there is no upper limit on gene number but we would not advise to search more than 10-15 genes.

		With default settings you will see read counts normalized by method applied in median-of-ratio method as described in Anders and Huber (DESeq/DESeq2) and divided by maximum value of expression level, so all values vary from 0 to 1.
		To change normalization type from Med to TMM (as in edgeR), click here.
		Each sample box is colored with graded purple color.
		To get read counts without normalization to maximum value of expression level, click here. You will see values of raw read counts normalized by method you chose before.
		Color of sample boxes is based on read counts value and varies from 0 to maximum expression value when read counts number type is set on “Raw Norm” or to 1 for “0-1” normalization. You can remove color information by clicking on “No” or set it to fold change-based coloring (this option works only when using differential gene expression analysis block, see below).
		Data for all samples can be saved as .png or downloaded in Excel format.

Differential gene expression.

This block allows to see whether there are the statistically significant differences in expression of a given gene between samples. You can choose a software used for differential expression detection(DESeq, DESeq2 and BaySeq)
here

To proceed to the analysis, click on a box of the sample which be used as “control” during analysis.

After that you will see fold changes of expression level in sample boxes. “Control” sample is indicated be the red frame and have fold change value 1. For samples that do not significantly differ from “control” fold change is “-“.

		Coloring of sample boxes (which is read counts-based by default) can be changed to fold change-based. At this case, values below 1, indicating down-regulation in this sample comparing to “control”, are colored by graded green, more intense for lower values (close to 0).
		Values above 1 (up to maximum fold change) are colored in orange, more intense for higher values.

Now you can go back to read counts display and still see coloring by fold change. “Control” sample is indicated as (basis - **).