Gene Expression Browser Instructions


About the Gene Expression Browser

This browser was designed using the R project Shiny platform ( To accurately deliver gene expression patterns, the interface requires an Ensembl gene ID as the input for a gene of interest. The Ensembl gene ID can be found using the Ensembl tool biomaRt, or the common gene name can be typed in the search window; possible Ensembl gene IDs (e.g., genes with similar names, paralogs) will show up at the top of the interface. Two graphs and one summary table will be generated after the Ensembl gene ID is typed into the input box

1. A bar graph showing gene expression in 9 organs;

2. Expression pattern for a gene over a 24-hour period in all organs;

3. A table showing expression pattern statistics (e.g., circadian expression pattern, peak expression time).

Example: Say you are interested in per gene family expression patterns but not sure how many “per” genes are in the X. maculatus genome, or not sure which “per” gene to study if there are paralogs within the fish genome. Type in “per” in Gene Name text box. The table on top right of the screen will display all genes in the Xiphophorus genome that contain “per”. If “per1b” is determined to be the gene of interest, type in the “per1b” Ensemble gene ID (ENSXMAG00000015314) into the Ensembl Gene ID text box. This process will trigger the generation of all graphs and tables.

About the data:

Mature (age ≥ 9 months) female and male X. maculatus (Jp 163A line in the 116th generation of inbreeding) were maintained on a 13 hr light/11 hr dark cycle under 10,000 K FL. Fish were fed with flake food at precisely 8 a.m. and brine shrimp at 3 p.m. during the synchronizing period. Circadian Time (Ct) 0 hr was assigned to 5 a.m. The onset of the light phase was Ct 2.5 hr and offset of the light phase, Ct 15.5 hr. Skin, brain, liver, gills, muscle, eyes, heart and gonads (ovaries for female and testis for male), were dissected from two female and two male X. maculatus Jp 163A at each Ct point (0, 3, 6, 9, 12, 15, 18 and 21 hr) for RNA isolation and subsequent identification of gene expression patterns.

The gene expression dataset was established using organ specific expression profiles over a 24-hr period by analyzing RNA-Seq (Sequencing library strategy: Illumina TruSeq; 150bp paired-end fragments) data. Short sequencing reads were filtered, mapped to X. maculatus genome v5.0 and gene expression quantified using Tophat2-featureCounts pipeline, facilitated by Ensembl gene annotation.

Gene expression was first normalized to Count Per Million Reads (cpm), followed with averaged cpm values calculated using biological replicates per time point. Relative expression at each time point was further calculated as relative expression = (Average cpm of biological replicate) / (Average cpm of all time points). Expressed genes of each organ were first identified by applying expression criteria to each gene of each organ-specific expression profile. A gene is considered expressed if its cpm value from at least two of the eight time points are larger than 1. Subsequently, a Rhythmicity Analysis Incorporating Nonparametric (RAIN) algorithm was applied to normalized and scale gene expression profiles to identify genes exhibiting oscillating circadian expression patterns. A p-value cutoff of < 0.01 was further used to identify genes that show oscillating expression patterns.