Record columns with their rows based on their names

[asddsaasddas]

I have a huge chunk of raw data ranging from column A to AM. The number of rows are not fixed and are based on the data retrieved from the system. Within columns A to AM, the first row will always be the header. With this being said, I want to be able to choose which columns and their rows below them based on the header's value and delete the rest of the columns. For example, in column A to AM there are 3 columns that i want to record. I want to be able to record the 3 columns based on the header's name and the rows below these 3 columns. Lets say the first row (header) in columns B, F and G are named "ID", "NAME" and "CONTACT" respectively. I want to be able to record these 3 columns and their rows and delete the all the other columns. All in all, I want to be able to choose which columns to be recorded based on the values of the first row (header) and delete all the other rows. Thus, in my example, I would be left with column A, B and C populated with "ID", "NAME" and "CONTACT" as the values for the first row. After sorting the columns based on their names, I want to be able to bold the header (first row) for these 3 columns left. Any help will be appericated

 

[Fluff]

Duplicate How to specifically pick entire few different columns from a huge set of data

Please do not post the same question multiple times. All clarifications, follow-ups, and bumps should be posted back to the original thread.
Per forum rules, posts of a duplicate nature will be locked or deleted (rule 12 here: Forum Rules).