Increased PCR screening capacity using a multi-replicate pooling scheme

Adrian Viehweger; Felix Kühnl; Christian Brandt; Brigitte König

a day. It is, therefore, desirable to maximize the number of samples that can be tested per reaction. 23 Various approaches have been proposed to do so in the context of SARS-CoV-2 RT-PCR testing. 2,3 One 24 problem with the traditional pooling approach, where several samples are collected and tested collectively, 25 is that the number of positive pools that require individual retesting increases rapidly with the number 26 of positive samples in the overall population, henceforth called "prevalence". A high prevalence renders 27 traditional pooling ineffective. To mitigate this, we propose to test samples in replicates and distribute 28 them across multiple pools. The resulting "pool address" can then be used to resolve samples in one pool, 29 given the information from other pools that contain a replicate. While some previous studies have taken a 30 similar approach implicitly, 2 it has neither been investigated systematically for more replicates than two, 31 nor is there any software that would generate and resolve the corresponding pooling layout for laboratory 32 use. 33 We therefore introduce "clonepool", a pooling framework to maximize the effective number of samples 34 per PCR reaction. "Effective" refers to the fact that samples in positive pools, whose status cannot be We tested the proposed clonepool algorithm using simulated data. We assumed no pipetting errors, 50 which can be achieved, e.g., through the use of a pipetting robot. We also assume that 94 pools are 51 available, which corresponds to a 96-well plate with two wells reserved for a positive and a negative 52 control. Furthermore, we assume that there are no false positive or false negative PCR reactions.