Lack of humility and a penchant for dogma are enemies of truth and fact, and all of us are susceptible to these human flaws, including scientists. When scientists hold fast to theories that are either not definitively testable, or that are not entirely consistent with observations or experimental results, then they are making a decision to ‘believe’ in their theory - irrespective of proof. This is, in fact, not different than religion. That's not the way science is supposed to work, but it often does.
The scientific method is not, unfortunately, applied nearly as frequently as it should be in experimental science. There are a lot of reasons for this, including conscious or unconscious bias in data interpretation in order to ensure publication. Recently a scientist who had been the head of cancer research at Amgen published a letter to Nature in which he described the efforts of a team of scientists from Amgen to replicate/confirm the results from 53 landmark scientific publications on cancer biology/research. They were only able to replicate 6 of the 53. http://www.nature.com/nature/journal/v483/n7391/full/483531a.htm
This doesn't mean that science can't be trusted. Science has been a big part of my life for a long time. It's a very important part of who I am. However, like any other endeavor that involves human beings, it is both fallible and corruptible.
The bottom line is that science doesn't ‘disprove’ God, although there are those who try to ascribe this function to science. I wholeheartedly stand by my assertion that belief in God and in the scientific method are in no way mutually exclusive, and that in searching for scientific truth one is often spiritually moved by what one finds.
Ping for later...
The technical terms are "methodological naturalism" versus "philosophical naturalism".
Science by definition is based on "methodological naturalism".
Some irreligious scientists claim that only "philosophical naturalism" is valid.
But in effect their "philosophical naturalism" simply fills the void left by the absense of a theistic component in their overall understandings.
And a basic problem with "philosophical naturalism" can bee seen in words variously attributed to Haldane or Eddington:
That "strangeness", or in Haldane's words, "queerness" suggests there may be natural limits to "methodological natrualism".