No negative control pair passes pairwise QC.

[AAAAAbetter]

Your package has been extremely helpful!

However, when running the code, I encountered the following error:
"Constructing negative control pairs.
Error in construct_negative_control_pairs_v2(sceptre_object = sceptre_object, :
No negative control pair passes pairwise QC."

My original code is as follows:

sceptre_object <- import_data_from_cellranger(
  directories = directories,
  moi = "high",
  grna_target_data_frame=grna_target_data_frame_highmoi
)

positive_control_pairs <- construct_positive_control_pairs(sceptre_object)

discovery_pairs<- construct_trans_pairs(
  sceptre_object = sceptre_object,
  positive_control_pairs = positive_control_pairs,
  pairs_to_exclude = "pairs_containing_pc_targets"
)

sceptre_object <- sceptre_object |>
  set_analysis_parameters(
    discovery_pairs = discovery_pairs,
    positive_control_pairs = positive_control_pairs,
    side = "left",
    grna_integration_strategy = "singleton",
    multiple_testing_alpha = 0.2
  ) |>
  assign_grnas(parallel = T,n_processors=2) |>
  run_qc(p_mito_threshold = 0.075) |>
  run_calibration_check(parallel = T, n_processors=2) |>
  run_power_check(parallel = T,n_processors=2) |>
  run_discovery_analysis(parallel = T, n_processors=2)

After adjusting grna_integration_strategy = "singleton" to the default option, the code runs without issues. Could you please explain the potential reasons for this error? What adjustments can be made during the experiment and sequencing to avoid this error?

[timothy-barry]

Good to hear!

The issue that you are encountering relates to pairwise QC. A negative control pair is analyzed only if it passes pairwise QC. You can tinker with the pairwise QC thresholds by setting n_nonzero_trt and n_nonzero_cntrl in set_analysis_parameters(). The default value of these parameters is 7; decreasing this value will make pairwise QC more liberal, thus increasing the chance that a negative control pair will pass pairwise QC.

If you don't mind, would you mind telling me a bit more about your experiment? In particular, how many negative control gRNAs do you have? How many targeting gRNAs do you have? Any how many total targets do you have? (In other words, how many targeting gRNAs per target site)?

[AAAAAbetter]

Thanks for your reply!
I have a total of 6000 gRNAs in my library, there are about 1200 genes in this library, an average of 5 sgRNAs per gene, and 31 gRNAs are non-targeted gRNAs.