Graduate schools are much less about GPA/GRE than med schools are about GPA/MCAT. Usually, there's a cutoff which may be a hard line or the personal opinion of whoever the first person to look at your application. If you're not over those cutoffs, no one's going to look at anything else. If you're over them, then you'll be considered. How far over them you are really isn't a big issue, although extremely high values will get some attention.
After you've crossed that hurdle, PhD admissions comes down to how confident the admissions committee is that you'll be productive once admitted. They want candidates they're confident will produce results when given a lab bench and reagents (or the equivalent for a non-bench program). If you don't have research, ultimately, all they can do is guess. And they're not going to commit funding to a candidate where guesswork is all they have to go on. ECs are good but not everything. Same with volunteer work. What it comes down to is recommendations, research, and publications. Research is the most important. I didn't have any publications when I applied for my PhD, but I had some cool research experience that they were really excited about. If you don't have research, expect a lot of programs to write a polite response asking you to do an MS and reapply in two years.