At the moment, https://github.com/JetBrains-Research/jbr/issues/250 makes a great job that allows user to find the required region by label & show it in the browser. The other popular usecase is to show some BEST / WORST / AVERAGE regions by some score. Typically user provides a BED file with several numeric columns (not only 'score' column from BED fromat notation). So it will be very useful to allow users to sort values by some selected column + view values in the regions list as well as labels and positions. Due to performance reasons, let's use only one (or severals and do not add selected)column all available numeric columns to the range list table. For simplicity, the initial implementation could allow only one column, e.g. just take current score column BED from bed-like track. So user could select score column for track using existing UI ('track contect menu') and range list will always use this column if it is available. Probably CGI track doesn't provided any score column and we could force it to use Observed to expected CpG
Use case example:
I have a BED like track with differentially methylated regions. Such track provided several metrics for each region. E.g. p-value [0..1] and methylation difference in range from [-1..1]
User selects p-value column as score column, opens range list, sort regions ascending by score column and checks in browser values with minimal pvalue, e.g. most significant regions
Then, the user sorts descending and checks several insignificant regions.
Then, the user scrolls the list and selects something near the pvalue 0.05, e.g. some boundary cases
Next user does the same but for methylation difference column
At the moment, https://github.com/JetBrains-Research/jbr/issues/250 makes a great job that allows user to find the required region by label & show it in the browser. The other popular usecase is to show some BEST / WORST / AVERAGE regions by some score. Typically user provides a BED file with several numeric columns (not only 'score' column from BED fromat notation). So it will be very useful to allow users to sort values by some selected column + view values in the regions list as well as labels and positions. Due to performance reasons, let's use only one (or severals and do not add selected)column all available numeric columns to the range list table. For simplicity, the initial implementation could allow only one column, e.g. just take current score column BED from bed-like track. So user could select score column for track using existing UI ('track contect menu') and range list will always use this column if it is available. Probably CGI track doesn't provided any score column and we could force it to use
Observed to expected CpGUse case example: