Gangan Prathap (National Institute of Science Communication and Information Resources, New Delhi) is the author of this preliminary study (pdf) ranking Indian universities based on papers listed in the Web of Science database. In order to do a fair comparison so that universities are compared only with similar institutions, he has wisely left out the institutions of national importance [Grrr, I hate this description --it implies universities are not of any "national importance"].
Here's the list of the top 20 research universities in India (based on their publications in the year 2008):
2. Banaras Hindu
8. Aligarh Muslim
10. CMC, Vellore
12. Jawaharlal Nehru
17. Sri Venkateswara
18. Guru Nanak Dev
One of the things he notes is the year-to-year variability in the rank order. For example, Mangalore University, with an overall University ranking of 20 in 2008, has an All India Rank of 23 in 2007 and 86 in 2004! This is clearly because the ranking is based on just one factor which, for any university, has a pretty large year-to-year variability.
Another observation from Pratap's data: the number of papers from the top 20 universities has registered a near 100 percent increase -- from 4034 to 8005 -- during the five year period of 2004-08.
The presence of CMC (Vellore) so high in that list (as opposed to better known, and better funded medical institutions such as AIIMS or PGI, Chandigarh) is also noteworthy. I think. [Update: As pointed out by Pratik in his comment, both AIIMS and PGI come under institutions of national importance, and therefore are left out by Pratap in this exercise.]
As Pratap points out, this is only the first such exercise, and it uses quantity -- the number of papers -- as its yardstick. He indicates that this effort may be expanded to include quality indicators such as citation counts.
[Speaking of which, I should link to this Current Science editorial by Prof. P. Balaram saying some nasty things about scientometrics.]
Anyways, if you want to know how the Indian Institutes of Technology (and IISc as well) have done in terms of publications during these last five years, check out Giridhar's post! [Update: Giridhar informs us In the comments section, Giridhar informs us that he has compiled a list (pdf) of top 40 Indian universities based not just on their publications, but also on citations and the h-index. See this post for the context.]