The guide answers an address question
A guide helps determine where the system binds, while its structural features help the editor operate. Different Cas proteins have different requirements. The guide does not independently transport the whole system into the desired organ.
The public assessment supports a high-level comparison of guide roles and platform assumptions. It is not a clinical guide-design protocol or a claim that any sequence is safe.
[C04] [C05] [C06] [C08] [C09] [C10] [C14]
Specificity is context-dependent
The Korean abasic-guide study described improved specificity in its experimental systems. That is a useful comparison candidate, not evidence of zero unwanted cleavage.
Single-cell and in vivo profiling highlighted that outcome patterns can vary across cells and organs. A bulk average may conceal events visible in smaller populations; a deliberately promiscuous experimental guide cannot supply a universal error rate.
[C04] [C05] [C06] [C08] [C09] [C10] [C14]
Delivery is another engineering problem
Compact Cas12m, virus-like particle delivery to primary myeloid cells and lipid-nanoparticle prime editing address different constraints. Their results should remain attached to the actual cell or disease-model setting.
A result in mouse liver does not establish delivery in aged brain or muscle. Platform size, cargo and cell context need to be considered only after a clear biological target and function have been selected.
[C04] [C05] [C06] [C08] [C09] [C10] [C14]
What to request from a partner
A partner assessment should identify the intended cell context, delivery evidence, measurement strategy, sample and data permissions, outcome definitions and known blind spots. Specificity, biological benefit and delivery efficiency should be recorded separately.
The practical output is a documented feasibility question. A public guide-generation interface cannot substitute for model-specific validation, genome-safety assessment or commercial-use rights.