Something like this should work:

\s+Income\s+\d\s+Wages,salaries,tips,etc.Attach\s+Form\(s\)W-2\s+7\s+([\d\.,\s]+)

Basically, you take the sentence and replace the spaces with \s (\s+ means one or more consecutive spaces). Then you add a \ in front of regular expression reserved characters like periods, parentheses and such. Then you surround the part you want extracted by parenthesis. Then you look at the data to guess the pattern. In this case, we have numbers (\d), commas, periods (\.) and spaces (\s). So we want anything in that range of characters.