It normally makes a lot of sense to link from the main domain to its subdomains. However, a lot here is about intention. Do the keyword subdomains exist for the visitor's benefit, or for your site's benefit on Google? Part of this answer would hang on how full or deep each subdomain is. For instance, would this subdomain be solid enough to stand on its own as a website?
Another, related question may be even more important. Do the LINKS to those subdomains exist for the visitor's benefit or for your site's benefit with Google? You say that you use footer links, and that sets off alarms for me - that the links are the kind that a visitor might not even notice.
In the site navigation one URL placed in robots.txt, throughout the website respective link will present, so we have added "nofollow" attribute as well.Is it correct idea?
It sounds odd to me to have a URL from the site navigation to be disallowed in robots.txt. Normally, main navigation point to the top level pages for the site, those that are important to notice. Sometimes a main navigation link points to a page that requires a log-in, but googlebot won't be able to log in anyway.
So why do you not want to see a page crawled that's in the main navigation?