Question 1

Should the screen be genome-wide or focused?

Accepted Answer

Genome-wide gives unbiased discovery but demands large cell numbers, deep sequencing and considerable capacity at every passage. A focused library covering one pathway or a druggable subset allows more replicates, more genotypes and better statistics for the same effort. Where the hypothesis is already narrow, focused libraries usually make a stronger thesis.

Question 2

Why does copy number cause false hits?

Accepted Answer

Because cutting at an amplified locus produces many breaks and a damage response that reduces fitness regardless of which gene was targeted. Amplified regions therefore look depleted without any real dependency. Correction methods exist, should be applied and named, and hits within amplified regions deserve extra caution.

Question 3

What comparison structure does a synthetic lethal claim need?

Accepted Answer

A between-genotype comparison. Depletion within a mutant line means little without the matched control, since common essential genes deplete everywhere. Isogenic pairs give the cleanest contrast, while panels of characterised lines allow the interaction to be tested against genotype across varied backgrounds. Either design should be stated explicitly.

Question 4

How should hits be validated?

Accepted Answer

With independent reagents and orthogonal logic: separate guides absent from the original library, a rescue construct restoring expression, and a chemical inhibitor where one exists. Validation should repeat the genotype comparison rather than only confirming depletion in the mutant, since that alone does not establish an interaction.

Question 5

What screen quality metrics should be reported?

Accepted Answer

Library representation after infection, replicate correlation, separation between known essential genes and non-targeting controls, and the depletion distribution across timepoints. These let a reader judge whether the screen had power to detect anything, and a screen without adequate control separation cannot support hit claims.

Question 6

How much does context dependence limit generalisation?

Accepted Answer

Considerably. Many interactions hold in one tissue, medium or genetic background and disappear in another, which is a genuine finding rather than a failure. Testing across more than one background, and reporting where the interaction is absent, is more useful and more defensible than presenting a single-line result as general.

Question 7

What sequencing and compute capacity is needed?

Accepted Answer

Enough sequencing depth for reliable guide counts at every timepoint and replicate, plus storage and pipeline capability for the counts and their analysis. Sequencing cost and scheduling belong in the plan before the screen starts, since insufficient depth cannot be repaired afterwards and repeating a genome-wide screen is rarely feasible.

Biotechnology

Biotechnology

Research Frontiers in Synthetic Lethal Interaction Screening