There is no point. Even if you can verify that the email address is syntactically valid, you’ll still need to check that it was not mistyped, and that it actually goes to the person you think it does. The only way to do that is to send them an email and have them click a link to verify.
Therefore, a most basic check (e.g. that they didn’t accidentally entered their street address) is usually enough. Something like: it has exactly one @ sign, and at least one . in the part after the @:
[^@]+@[^@]+\.[^@]+
You’d probably also want to disallow whitespace — there are probably valid email addresses with whitespace in them, but I’ve never seen one, so the odds of this being a user error are on your side.
If you want the full check, have a look at this question.
Update: Here’s how you could use any such regex:
import re
if not re.match(r"... regex here ...", email):
# whatever
Python ≥3.4 has re.fullmatch which is preferable to re.match.
Note the r in front of the string; this way, you won’t need to escape things twice.
If you have a large number of regexes to check, it might be faster to compile the regex first:
import re
EMAIL_REGEX = re.compile(r"... regex here ...")
if not EMAIL_REGEX.match(email):
# whatever
Another option is to use the validate_email package, which actually contacts the SMTP server to verify that the address exists. This still doesn’t guarantee that it belongs to the right person, though.