Your best information will come from a series of tests.
You could reference the 2005 NDS, Section 11 for more information - you will need to know the specific gravity of the wood. For a #6 wood screw, the values vary from 38 to 209 pounds per inch of penetration into the side grain, multiplied by the appropriate adjustment factors in Table 10.3.1. Your final calculated values from the NDS will likely be substantially less than your test results. But then, your test results will also vary dramatically depending upon the nature of the grain that the screws are driven into.
None of the above considers the tensile strength of the screw itself.
I can't comment on NDS but the wood standard I use gives a withdrawal capacity of 170 lbs per inch of penetration for a No. 6 screw. It will vary depending on the quality of the wood. The thread doesn't come into the calculations for withdrawal from wood.
Drywall screws are not wood screws so the withdrawal values will probably not be the same. Without a ICC report on the withdrawal values of the drywall screws, I would use the nail values, knowing that the actual value would be higher.
As for testing, you will need to verify that the wood you are testing in meets the actual sg for that species as 95% of that species should be higher than the species sg. This is because of the 5% rule for determining the lumber values under the grading rules.
Agree with Woodman, Drywall Screws are NOT wood screws. They are also very brittle, if you have any type of cyclic loading, the use of drywall screws is not realistic and should not be considered as a structural connector as they are not intended or tested to be such.
