Assignment review
Switch between recorded results. No reads are reanalysed and no policy is recommended.
Read outcomes
All three policies use the same input reads and extraction window.
Guide counts
Select a guide to inspect its recorded counts and available read evidence.
Printed tables contain the currently displayed page only. Export the filtered TSV to retain every matching guide.
Read-state transitions
Select a cell to inspect a read-state transition.
The diagonal means the same state, not necessarily the same target. A transition matrix alone cannot establish identical read membership.
Provenance
Input identities, artifact checks and analysis settings.
Artifact checks
Per-guide totals match unique outcomes. Transition margins account for all reads. These checks do not authenticate the report or validate the assay.
Input files
Input files are not embedded or re-read. A recorded checksum is not proof of the biological sample, source authenticity or appropriateness.
Matching rules and interpretation limits
Exact
Exactly one target at distance zero. Duplicate target sequences remain ambiguous.
Radius k=1
Exactly one target anywhere within one substitution. Even an exact hit is ambiguous when another target is in that radius.
Best k=1
A unique nearest target within one substitution. Equally near targets remain ambiguous.
Quality characters are validated, not used to weight assignments. N and IUPAC symbols are literal, not wildcards. No indels, offset inference, reverse-complement search, cell/UMI processing or downstream screen statistics. Unique is an assignment outcome, not proof of biological origin.