Generally speaking, you are always looking for "prior art". This means, in principle, ANY information PUBLICLY AVAILABLE at the time of your patent application.

If you say that someone has "submitted, but not published anywhere", then -most likely-the information is not publicly available and thus shouldn't block your patent application.

The case may, however, get a little more complicated if the other side can prove they did a similar thing before (e.g. by a dated/stamped/signed lab journal). In some countries the rule is "first to file", in others "first to invent".  You have to see which is the case for you.

It will depend on the interpretation of prior art in the specific jurisdiction that you want to patent your invention. Under the US patent law even if it is publicly known(even though it is not published) it can still be prior art.

That said, if your claims point to specific aspects of the invention that are not contemplated in the disclosure it can still be accepted. It all comes down on how you specify your claim and how it differentiates from the prior art.

Disclaimer: I am not a patent lawyer, please consult with one, especially on this matter.

There are no countries today that have a "first to invent" patent system.  "First to file" is now the universal rule.

A thesis, if it's in a university library AND was properly indexed before you filed your application, is considered "published" for purposes of patent law.  The key factor is actually the indexing, which is what makes the thesis available to the public.  Even if the only way to find it is to physically go to that one particular library and pick through the card catalog, the document will be regarded as "published." (This is US law, and the definition of "available to the public" may vary from country to country.)

The patent application is still good - you have to wait for a patent examiner to look at the thesis, and decide if it makes the invention already known or obvious, and not patentable for that reason.  This assumes, of course, that the examiner knows about the thesis.  In some countries (the US in particular), you're obliged to tell the patent office about all relevant publications, so if you've seen the thesis, you have to give a copy to the examiner.  Failure to do that can cause any patent you get to be declared invalid.  (If you haven't actually seen it, however, there's no obligation.)

http://www.uspto.gov/web/offices/pac/mpep/s2128.html

@ James. Regarding "first to file": I was aware of the changes in the US, but not 100% sure if this is/was the case all over the world. Makes things easier if this is "now the universal rule". Thanks for the clarification.

thank you all