It’s challenging. There’s not a lot of good information out there, and made worse by polyhybridization, which creates unstable results from seed. I have started

It’s challenging. There’s not a lot of good information out there, and made worse by polyhybridization, which creates unstable results from seed. I have started using thin layer chromatography kits on occasion, which provide some level of cannabinoid testing. In general, I would recommend getting outside of mainstream strains and looking into heirlooms and landraces. Strains I’ve found to be “low potency” are either low cannabinoid (not much resin content) or hemp (little to no THC). Strains that have not been farmed and are collected from wild populations are difficult to grow and are probably only for people who are curious about the genetic extremes of the plant. Inbred lines originating from hash cultures are a good place to start. American heirlooms from the 60s-80s are another good option. In the US, “CSI Humboldt” offers a number of options, or “authentic genetics”. For the more adventurous, the real seed company collects many heirlooms and wild accessions; however they are based outside the US. As a general recipe, if you grow a cbd hemp strain and pollinate it with an thc monocannabinoid strain, then the first generation will yield roughly equal parts THC and CBD, while inbreeding the line will create substantial variation that needs to be stabilized through selection. Virtually all of the marketed cannabis has been repeatedly outcrossed, and the seed lines are highly variable (except in cannabinoids because of repeated THC selection). So commercial strains are selected from clones and mother plants are maintained to create a uniform product.