False negative rates in Drosophila cell-based RNAi screens: a case study

BMC Genomics

Table 3 Model of RNAi reagent disambiguation methods under one, two or three reagents per gene.

	1 RNAi Reagent/Gene	2 RNAi Reagents/Gene	3 RNAi Reagents/Gene
"Lenient" Rule A: Number of False Negatives	R_FN × H	$R_{F N}^{2} \times H$	$R_{F N}^{3} \times H$
"Lenient" Rule A: Number of False Positives	R_FP × N	$(R_{F P}^{} \times 2 - R_{F P}^{2}) \times N$	$(R_{F P}^{} \times 3 - R_{F P}^{2} \times 3 + R_{F P}^{3}) \times N$
"Stringent" Rule B: Number of False Negatives	R_FN × H	$(R_{F N} \times 2 - R_{F N}^{2}) \times H$	$(R_{F N} \times 3 - R_{F N}^{2} \times 3 + R_{F N}^{3}) \times H$
"Stringent" Rule B: Number of False Positives	R_FP × N	$R_{F P}^{2} \times N$	$R_{F P}^{3} \times N$
"Balanced" Rule C: Number of False Negatives	R_FN × H	$(R_{F N} \times 2 - R_{F N}^{2}) \times H$	$(R_{F N}^{2} \times 3 - R_{F N}^{3} \times 2) \times H$
"Balanced" Rule C: Number of False Positives	R_FP × N	$R_{F P}^{2} \times N$	$(R_{F P}^{2} \times 3 - R_{F P}^{3} \times 2) \times N$

N = The number of genes in the screening library, H = The number of genes a screen should uncover under ideal conditions, R_FN = Fraction of reagents of H that fail. R_FP = Fraction of reagents of N that are false positives. Rule "A": Only one reagent targeting a gene need to be a "hit" for the gene to be called a "hit". Rule "B": All reagents targeting a gene need to be a "hit". Rule "C": More than half of reagents must score as a "hit".

ISSN: 1471-2164