Pooling Approach Research Articles

A novel fuzzy pooling based modified ThinNet architecture for lemon fruit classification

Computer vision functions like object detection, image segmentation, and image classification were recently getting advance due to Convolutional Neural Networks (CNNs). In the food and agricultural industries, image classification plays a critical role in quality control. CNNs are made up of layers that alternate between convolutional, nonlinearity, and feature pooling. In this article, we proposed Fuzzy Pooling, a novel pooling approach that works based on fuzzy logic, that can increase the accuracy of the CNNs by replacing the conventional pooling layer. This proposed Fuzzy Pooling was put to the test with CIFAR-10 and SVHN data sets on single layer CNN, and it outperformed previous pooling strategies by achieving 92% and 97% classification accuracy. This proposed Fuzzy Pooling layer was replaced the Max Pooling layer in the ThinNet architecture, and it was trained using the back propagation method. It was demonstrated experimentally on the Lemon fruit data set to classify the fruits into three categories such as Good, Medium, and Poor. In order to classify the lemon fruit into three categories, the 1000 Fully Connected layer in ThinNet architecture was replaced with three Fully Connected layers. The Modified ThinNet architecture called ThinNet_FP was trained with a learning rate of 0.001 and achieved 97% accuracy in classifying the images and outperformed previous CNN architectures when trained on the same data set.

Role of financial sector development in foreign direct investment inflows in BRICS

This study examined the influence of financial sector development on FDI inflows in BRICS using panel data (1991–2020) analysis methods. The influence of the complementarity between the financial sector and human capital development on FDI inflows was also examined in the context of BRICS using the same data set and econometric methodologies. The advantage of this study is that the results are used as a basis by BRICS countries to develop financial sector development policies that attract significant FDI inflows. Financial sector development (model 2 and 3 of the pooled ordinary least squares approach) significantly enhanced FDI inflows. Human capital development (model 3 of the fully modified ordinary least squares) was found to have had a significant positive effect on FDI inflows in BRICS group of countries. The combination between financial and human capital development under (1) model 1 of the fully modified ordinary least squares and (2) models 2 and 3 of the pooled ordinary least squares (POLS) was observed to have significantly enhanced FDI inflows in BRICS. The study outlines the financial and human capital development recommendations that need to be implemented to facilitate more FDI inflows. Acknowledgment Kunofiwa Tsaurai is grateful for the moral support from his employer, University of South Africa.

Staff pooling in healthcare systems – results from a mixed-methods study

ABSTRACT This study examines how staff pooling can be used to create a higher service level at a predetermined total capacity in the healthcare sector. We develop new empirical knowledge through a systematic empirical study, using a mixed-methods approach, with a preliminary interview study followed by a principal quantitative survey study, with data from a multihospital system. The purpose was to explore practical barriers for a staff pooling strategy in healthcare systems. Three barriers were identified:recruitment difficulties, community view, and specialisation. Significant differences in perceived height among these barriers were found. The results from this study have important managerial implications for healthcare systems when implementing a staff pooling approach. This study contributes to the existing literature since, to the best of our knowledge, no previous research has been done where barriers to staff pools are systematically identified using a holistic approach that includes all healthcare professions in a multihospital system.

Forecast combination-based forecast reconciliation: Insights and extensions

In this paper, we build upon a recently proposed forecast combination-based approach to the reconciliation of a simple hierarchy (Hollyman R., Petropoulos F., Tipping M.E., Understanding forecast reconciliation, European Journal of Operational Research, 2021, 294, 149–160) and extend it in some new directions. In particular, we provide insights into the nature and mathematical derivation of the level-l conditional coherent (LlCC) point forecast reconciliation procedure for an elementary two-level hierarchy. We show that: (i) the LlCC procedure is the result of a linearly constrained minimization of a quadratic loss function, with an exogenous constraint given by the base forecast of the top level series of the hierarchy, which is not revised; and (ii) endogenous constraints may also be considered in the same framework, thereby resulting in level conditional reconciled forecasts where both the top and the bottom level time series forecasts are coherently revised. In addition, we show that the LlCC procedure (i.e., with exogenous constraints but the result also holds in the endogenous case) does not guarantee the non-negativity of the reconciled forecasts, which can be an issue in cases when non-negativity is a natural attribute of the variables that need to be forecast (e.g., sales and tourism flows). Finally, we consider two forecasting experiments to evaluate the performance of various cross-sectional forecast combination-based point forecast reconciliation procedures (vis-à-vis the state-of-the-art procedures) in a fair setting. In this framework, due to the crucial role played by the (possibly different) models used to compute the base forecasts, we re-interpret the combined conditional coherent reconciliation procedure (CCCH) of Hollyman et al. (2021) as a forecast pooling approach, and show that accuracy improvements may be obtained by adopting a simple forecast averaging strategy.

Characterization of glyphosate and AMPA concentrations in the urine of Australian and New Zealand populations

Glyphosate is the most used herbicide globally, but our understanding of human exposure and how different uses affect exposure is not well understood. The aim of this study was to obtain the first data on glyphosate and its primary degradation product aminomethylphosphonic acid (AMPA) concentrations in pooled and individual urine from the Australia and New Zealand region using a sensitive direct injection method and compare results with studies from elsewhere. Pooled urine samples from the Australian general population (n = 125 pools representing >1875 individuals) and individual urine samples (n = 27) from occupationally exposed New Zealand farmers were analysed by LC-MS/MS. Glyphosate was detected above the LOD (0.20–1.25 μg/L) in 8 % of the Australian population pooled urine samples with most detections in the 45–60 years age group. Furthermore, glyphosate (0.85 to 153 μg/L) and AMPA (0.50 to 3.35 μg/L) were detected in 96 % and 33 % of farmers, respectively. The maximum glyphosate urine concentration was 1.7 times above the recommended acceptable daily intake (ADI), when assuming a urinary excretion rate of 1 %. The pooled sampling and analysis approach proved effective for rapid large-scale screening of populations and could be used to determine where targeted and more specific individual sampling may be required.

Stereo Selective Total Synthesis of 4‐(S)‐Hydroxy‐5‐(R)‐octyldihydrofuran‐2‐one: Harvestmen Lactone

AbstractA facile and stereoselective synthesis of Harvestmen lactone 1 was accomplished in a linear synthetic approach over 14 steps with an overall yield of 10.1 %. This total synthesis was carried out using a chiral pool approach from commercially available D‐glucose. The key reactions employed in this synthetic approach were Wittig reaction, Barton‐McCombie Deoxygenation and 2,2,6,6‐Tetramethylpiperidin‐1‐yl)oxyl (TEMPO) mediated radical oxidation.

Encoding Cardiopulmonary Exercise Testing Time Series as Images for Classification using Convolutional Neural Network.

Exercise testing has been available for more than a half-century and is a remarkably versatile tool for diagnostic and prognostic information of patients for a range of diseases, especially cardiovascular and pulmonary. With rapid advancements in technology, wearables, and learning algorithm in the last decade, its scope has evolved. Specifically, Cardiopulmonary exercise testing (CPX) is one of the most commonly used laboratory tests for objective evaluation of exercise capacity and performance levels in patients. CPX provides a non-invasive, integrative assessment of the pulmonary, cardiovascular, and skeletal muscle systems involving the measurement of gas exchanges. However, its assessment is challenging, requiring the individual to process multiple time series data points, leading to simplification to peak values and slopes. But this simplification can discard the valuable trend information present in these time series. In this work, we encode the time series as images using the Gramian Angular Field and Markov Transition Field and use it with a convolutional neural network and attention pooling approach for the classification of heart failure and metabolic syndrome patients. Using GradCAMs, we highlight the discriminative features identified by the model. Clinical relevance- The proposed framework can process multivariate exercise testing time-series data and accurately predict cardiovascular diseases. Interpretable Grad-CAMs can be obtained to explain the prediction.

Orthographic Pooling: Learned Maximum Intensity Projection for Vertebrae Labelling.

Maximum intensity projection (MIP) is a standard volume-rendering technique for 3D volumetric data processing. For example, given a 3D CT data, it simply projects the voxel values with its maximum intensity on a specific view to output a 2D image. Recently, MIP is further combined with Btrfly Net for vertebrae labelling task. However, this simple reformations of 3D data leads to loss of rich context information in volumetric data. In this paper, we propose a learned orthographic pooling approach instead of image processing based MIP. Typically, a simple conv-simple and bottleneck pooling modules are introduced to learn the orthographic projection of 3D data and output 2D intermediate feature maps. To this end, the learned orthographic pooling helps preserve detail information of 3D context during projection. Furthermore, an unified Btrfly Net is provided for vertebrae labelling by integrating the orthographic pooling sub-network. The novel Btrfly Net with orthographic pooling sub-network is evaluated on the 2014 MICCAI vertebra localization challenge dataset. Compared to original Butfly Net with MIP, orthographic pooling, the learned MIP largely boosts the performance of vertebrae labelling.

Surveillance testing using salivary RT-PCR for SARS-CoV-2 in managed quarantine facilities in Australia: A laboratory validation and implementation study

SummaryBackgroundRegular repeat surveillance testing is a strategy to identify asymptomatic individuals with SARS-CoV-2 infections in high-risk work settings to prevent onward community transmission. Saliva sampling is less invasive compared to nasal/oropharyngeal sampling, thus making it suitable for regular testing. In this multi-centre evaluation, we aimed to validate RT-PCR using salivary swab testing of SARS-CoV-2 for large-scale surveillance testing and assess implementation amongst staff working in the hotel quarantine system in Victoria, Australia.MethodsA multi-centre laboratory evaluation study was conducted to systematically validate the in vitro and clinical performance of salivary swab RT-PCR for implementation of SARS-CoV-2 surveillance testing. Analytical sensitivity for multiple RT-PCR platforms was assessed using a dilution series of known SARS-CoV-2 viral loads, and assay specificity was examined using a panel of viral pathogens other than SARS-CoV-2. In addition, we tested capacity for large-scale saliva testing using a four-sample pooling approach, where positive pools were subsequently decoupled and retested. Regular, frequent self-collected saliva swab RT-PCR testing was implemented for staff across fourteen quarantine hotels. Samples were tested at three diagnostic laboratories validated in this study, and results were provided back to staff in real-time.FindingsThe agreement of self-collected saliva swabs for RT-PCR was 84.5% (95% CI 68.6 to 93.8) compared to RT-PCR using nasal/oropharyngeal swab samples collected by a healthcare practitioner, when saliva samples were collected within seven days of symptom onset. Between 7th December 2020 and 17th December 2021, almost 500,000 RT-PCR tests were performed on saliva swabs self-collected by 102 staff working in quarantine hotels in Melbourne. Of these, 20 positive saliva swabs were produced by 13 staff (0.004%). The majority of staff that tested positive occurred during periods of community transmission of the SARS-CoV-2 Delta variant.InterpretationSalivary RT-PCR had an acceptable level of agreement compared to standard nasal/oropharyngeal swab RT-PCR within early symptom onset. The scalability, tolerability and ease of self-collection highlights utility for frequent or repeated testing in high-risk settings, such as quarantine or healthcare environments where regular monitoring of staff is critical for public health, and protection of vulnerable populations.FundingThis work was funded by the Victorian Department of Health.

Accessing Diverse Cross-Benzoin and α-Siloxy Ketone Products via Acyl Substitution Chemistry.

An approach to diverse cross-benzoin and α-siloxy ketone products which leverages a simple yet underutilized C-C bond disconnection strategy is reported. Acyl substitution of readily accessible α-siloxy Weinreb amides with organolithium compounds enables access to a broad scope of aryl, heteroaryl, alkyl, alkenyl, and alkynyl derivatives. Enantiopure benzoins can be accessed via a chiral pool approach, and the utility of accessible cross-benzoins and α-siloxy ketones is highlighted in a suite of downstream synthetic applications.

Exploring the Parametric Impact on a Deep Learning Model and proposal of a 2-Branch CNN for Diabetic Retinopathy Classification with Case Study in IoT-Blockchain based Smart Healthcare System

Smart healthcare has changed the way how the patient interacts with the specialists for treatment. However, security and support for various diseases are still the concern for such smart automated systems. One of the critical diseases namely Diabetic Retinopathy (DR), is a major concern for the person with prolonged diabetes and may lead to complete blindness irrespective of age groups. Moreover, in recent years blockchain has gained popularity in providing secure communication between sender and receiver. Hence, this work focus on designing a blockchain-based smart healthcare system for the early detection of diabetic retinopathy. However, early detection of DR impose complexities and requires expert diagnosis, which is not available everywhere. Hence, the proposed smart healthcare model contains a Computer-Aided Diagnosis (CAD) assistance for early detection of symptoms of the disease. The CAD model may assist the ophthalmologists in the early detection of DR, which requires intensive research in developing an efficient and accurate model that can operate without human interaction. Convolutional Neural Network (CNN) is one of the learning models in AI that has shown its potential in computer vision applications. However, the performance of CNN depends on several factors like the pooling approach, activation function, learning rate and CNN architecture. This study provides an empirical analysis of these factors to design the best model for early detection of DR. The best model can be used to develop IoT based smart devices to detect DR in diabetic patients. The study also explains the importance of IoT and blockchain-based technology for the development of smart healthcare systems. The values of the parameters and type of hyperparameters chosen from the study is used in a proposed 2-branch CNN model, and the model is validated using the Kaggle fundus image set. Analysis of various parameters and using their best values gives an outstanding performance in the proposed 2-branch CNN model.

Abstract 47: Identifying genetic vulnerabilities of chromosome 4p large copy number variants in triple negative breast cancer

Abstract Triple Negative Breast Cancer (TNBC) is characterized by the absence of common oncogenic drivers, limiting its treatment options; however, it exhibits recurrent large chromosomal deletions. We previously showed that chromosome 4p (chr4p) loss is a frequently observed large copy number variant in TNBC and is associated with poor prognosis. We also showed that chr4p deletion is an early event in tumor evolution and confers on cells a proliferative advantage. Here, we set out to uncover the genetic vulnerabilities associated with chr4p deletion in TNBC to identify novel therapeutic avenues for TNBC and enhance our understanding of the genetic mechanisms that maintain chr4p deletion in the genome. Whole genome sequence analysis of our TNBC Primary Tumor(PT)/Patient-Derived Xenograft (PDX) panel identified samples with copy neutral and deletion status of chr4p. These deletion regions span a large fraction of the chr4p arm. RNAseq analysis revealed that chr4p deletion is functionally significant since gene expression of ~80% of genes was reduced upon chr4p deletion. Chr4p deletion was associated with global transcriptomic changes and differentially expressed genes were enriched for proliferation, DNA replication, cell migration, activation of the innate immune response and protein translation. PDX-derived cell models from these samples showed lentiviral infectivity based on a control lentivirus expressing GFP. Additionally, these PDX cell models were shown to be editable using CRISPR-Cas9 through the targeting of core essential genes. We will use a pooled CRISPR-Cas9 approach to systematically screen for genetic vulnerabilities in TNBC PDX-derived cell models harbouring chr4p copy neutral or deletion state. To further investigate the genetic mechanisms buffering chr4p loss, we have leveraged publicly-available CRISPR-Cas9 genome-wide genetic screen data from the Cancer DepMap to identify putative synthetic lethal (pSL) partners with chr4p deletion in TNBC. This was accomplished by developing regression models and integrating them with results from established methods such as drugZ and MAGeCK. We identified pSL partners for chr4p at gene, segmental and arm levels. pSLs were enriched for mitochondrial, protein translation and proliferation pathways. We will validate the top candidates from these analyses in our cohort of chr4p deletion and chr4p copy neutral TNBC PDX-derived cell models and integrate them with the pooled CRISPR screens. Together, this work aims to reveal potential TNBC-specific therapeutic avenues for precision oncology. Citation Format: Michael Schwartz, Rohan Dandage, Lynn Karam, Alain Pacis, Hellen Kuasne, Anne-Marie Fortier, Sidong Huang, Guillaume Bourque, Traver Hart, Elena Kuzmin, Morag Park. Identifying genetic vulnerabilities of chromosome 4p large copy number variants in triple negative breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 47.

Abstract 2278: Expanding cancer therapy options by leveraging synthetic lethal interactions between druggable paralogs

Abstract Synthetic lethal therapies are a promising approach to expand therapeutic options for cancer patients. Since synthetic lethal therapies target tumor cells specifically, they have fewer off-target effects than oncogene targeted approaches. To identify new cancer drug targets, we focused on paralogs, ancestrally-duplicated genes that frequently retain redundant or overlapping functions. To find synthetic lethal human paralogs, we developed paired guide RNAs for Paralog gENetic interaction mapping (pgPEN), a pooled CRISPR-Cas9 single and double knockout approach targeting over 2,000 paralogs. We applied pgPEN to two cancer cell lines and found that 12% of human paralogs exhibit synthetic lethality in at least one context. To our knowledge, pgPEN represents the largest experimental assessment of human paralog synthetic lethality to date. We next identified paralog pairs to prioritize for follow-up study. A key drawback to synthetic lethal therapies is that many genetic interactions are context-dependent. To identify likely penetrant interactions, we compared our data to other published paralog screens and computational predictions of paralog synthetic lethality. We found that over 75% (n=96) of pgPEN hits were predicted to be broadly synthetic lethal, and nearly 10% (n=10) of pairs were synthetic lethal in multiple paralog screens. Finally, we prioritized paralogs targeted by existing small molecule therapies. Of the 122 synthetic lethal pairs identified by pgPEN, 16% (n=20) are currently druggable. We mined drug repurposing data from DepMap to find pairs where a paralog-targeting drug showed a stronger effect in cell lines with low target gene copy number or expression relative to cell lines with normal target gene copy number or expression. These drugs could selectively target cancer cells in cases where one or both paralogs is lost in tumors but retained in normal tissue. We also leveraged The Cancer Genome Atlas tumor sequencing data to find paralog pairs where one member is recurrently lost in cancer. Taken together, these studies identify druggable, highly penetrant synthetic lethal paralog interactions. In combination with tumor sequencing data, we identify high-priority paralog drug targets that can be further tested and translated to the clinic. Paralog synthetic lethal therapies could provide a relatively low-toxicity therapeutic approach to improve the efficacy of cancer treatments and prevent the emergence of acquired resistance. Citation Format: Phoebe C. Parrish, James D. Thomas, Austin M. Gabel, Shriya Kamlapurkar, Robert K. Bradley, Alice H. Berger. Expanding cancer therapy options by leveraging synthetic lethal interactions between druggable paralogs [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2278.

NiPA: Nicknames Pool Approach for Protecting Privacy of User Data in Intelligent Transportation Systems

The intelligent transportation system has made a huge leap in the level of human services, which has had a positive impact on the quality of life of users. On the other hand, these services are becoming a new source of risk due to the use of data collected from vehicles, on which intelligent systems rely to create automatic contextual adaptation. Most of the popular privacy protection methods, such as Dummy and obfuscation, cannot be used with many services because of their impact on the accuracy of the service provided itself, they depend on changing the number of vehicles or their physical locations. This research presents a new approach based on the shuffling Nicknames of vehicles. It fully maintains the quality of the service and prevents tracking users permanently, penetrating their privacy, revealing their whereabouts, or discovering additional details about the nature of their behavior and movements. Our approach is based on creating a central Nicknames Pool in the cloud as well as distributed sub-pools in fog nodes to avoid intelligent delays and overloading of the central architecture. Finally, we will prove by simulation and discussion by examples the superiority of the proposed approach and its ability to adapt to new services and provide an effective level of protection. In the comparison, we will rely on the well-known privacy criteria: Entropy, Ubiquity, and Performance.

Identifying and unlocking the value from heat decarbonisation in the United Kingdom

Decarbonisation of heat is critical to the UK achieving its net zero emissions target by 2050. With the low rate of asset turn-over in the heat sector, decarbonisation needs to start immediately. However, the risk and uncertainty regarding the most ‘cost optimal pathway’ across the range of technology options and the lack of a holistic UK heat policy is failing to provide a clear directive as to how market actors should address UK heat decarbonisation. This research applies a novel commercial perspective to provide additional insight beyond that which traditional cost optimal tools are able to offer. The value pool approach is used to determine the magnitude of economic opportunities and map their resilience across multiple net zero scenarios in decadal timesteps to 2050. Our work indicates that by 2050 an annual value of £28 billion potentially exists in decarbonising heat in the UK. Realising this value, however, is subject to significant path-dependency. Unlocking the value will require substantial additional policy and regulatory support, business model innovation and unprecedented levels of consumer engagement and protection in the history of the energy sector.

An asymptotic perspective on risk pooling: Limitations and relationship to transshipments

In this article we provide a novel perspective on risk pooling approaches by characterizing and comparing their asymptotic performance, highlighting the conditions under which one approach dominates the other. More specifically, we determine the inventory policy and the expected total costs of systems under physical and information pooling as the number of locations grows. We show that physical pooling dominates information pooling in settings with no additional per-location costs for operating the centralized system. In the presence of such costs, however, information pooling becomes a viable alternative to physical pooling. Through asymptotic analysis, we also address the grouping problem, the division of a given set of non-identical locations into an ordered collection of mutually exclusive and collectively exhaustive subsets of predetermined sizes and demonstrate that homogeneous groups, comprising locations with similar demand volatility, achieve a lower expected total cost. Finally, the convergence of the expected total costs and the base stock levels under the two pooling approaches is demonstrated through a simple numerical illustration. Our analysis supports the assertion that it is important to consider not only the individual characteristics of each location in isolation, but also the interactions among them, when designing pooling systems.

Testing the Pollution Haven Hypothesis with the Role of Foreign Direct Investments and Total Energy Consumption

The main objective of this study was to examine the nonlinear relationship between environmental deterioration and foreign direct investment for subpanels based on the country’s income level. In this study, the model’s determinants were total consumption of energy and electricity consumption, the share of renewable energy, and economic growth. Due to the observation of cross-sectional dependence, utilization of cointegration tests and panel data unit root were incorporated, which confirmed a mixed integration order. For the compliance of long-run and short-run relationships among the variables, a pooled mean group estimator panel auto-regressive distributed lag approach was incorporated. The results of long-run development support the pollution haven hypothesis; hence, ecological footprint is increased by the activities related to foreign direct investments. The obtained findings depend on the different subpanels based on the income level of countries. For the assurance of economic development sustainability in the energy sector, along with the electrical energy sector, customized policymaking is suggested by this study based on the particulars of each subpanel.

Does Down syndrome influence the outcomes of congenital cardiac surgery? A systematic review and meta-analysis.

Congenital cardiac surgery for individuals with Down syndrome (DS) has historically occurred at a reduced frequency. Little data are available regarding long-term post-congenital cardiac surgical outcomes. Limited sample sizes and clinical heterogeneity require a pooled analysis approach. To compare long-term outcomes post-congenital heart surgery between adults with and without DS. Databases (Medline, Embase, and PubMed) were searched utilizing terms related to DS and congenital heart disease.Studies that enrolled adults (>18 years) with operated congenital heart disease and compared long-term outcomes with respect to DS presence were included. All study designs were included, but those with limited/peri-operative follow-up, non-English texts, case studies, and literature reviews were excluded.Blinded screening, data extraction, and quality assessment were independently conducted by two reviewers. QUIPS criteria were used for risk of bias analysis. Both random- and fixed-effects models were used for meta-analysis. A total of 23 studies (n=10466) were included. Risk of bias was frequently high due to unblinded retrospective study designs and analyses limited in adjustment for other prognostic factors.Meta-analysis demonstrated no effect of DS on long-term mortality [hazard ratio (HR) 0.86, 95% confidence interval (95% CI) 0.6-1.23], to a maximum described follow-up of 38 years. Lower cardiac reoperation risk (HR 0.6, 95% CI 0.46-0.78) for individuals with DS was found on pooled analysis. Meta-analysis was limited by between-study variation. DS does not affect post-congenital cardiac surgical survival in adulthood. Reduced reoperation may reflect challenges in assessing functional and symptomatic status and/or concerns regarding perceived reoperation difficulties or likely benefits.

Boosting SARS-CoV-2 detection combining pooling and multiplex strategies

RT-qPCR is the gold standard technique available for SARS-CoV-2 detection. However, the long test run time and costs associated with this type of molecular testing are a challenge in a pandemic scenario. Due to high testing demand, especially for monitoring highly vaccinated populations facing the emergence of new SARS-CoV-2 variants, strategies that allow the increase in testing capacity and cost savings are needed. We evaluated a RT-qPCR pooling strategy either as a simplex and multiplex assay, as well as performed in-silico statistical modeling analysis validated with specimen samples obtained from a mass testing program of Industry Federation of the State of Rio de Janeiro (Brazil). Although the sensitivity reduction in samples pooled with 32 individuals in a simplex assay was observed, the high-test sensitivity was maintained even when 16 and 8 samples were pooled. This data was validated with the results obtained in our mass testing program with a cost saving of 51.5% already considering the expenditures with pool sampling that were analyzed individually. We also demonstrated that the pooling approach using 4 or 8 samples tested with a triplex combination in RT-qPCR is feasible to be applied without sensitivity loss, mainly combining Nucleocapsid (N) and Envelope (E) gene targets. Our data shows that the combination of pooling in a RT-qPCR multiplex assay could strongly contribute to mass testing programs with high-cost savings and low-reagent consumption while maintaining test sensitivity. In addition, the test capacity is predicted to be considerably increased which is fundamental for the control of the virus spread in the actual pandemic scenario.

Abstract B014: Massively parallel functional assessment of label-free mutant pools is a universal approach to parametrize mechanistic models of drug resistance evolution

Abstract Systematically scanning the effects of amino acid substitutions across a protein is a powerful technique in protein engineering, synthetic biology, and the study of drug resistance. In general, these selection experiments increase growth rates to select for mutants. Current methods use molecular barcodes to circumvent the resolution barriers imposed by conventional NGS sequencing of point mutations in mutant pools. Barcodes introduce synonymous mutations in the genome and assume neutrality, but it is now well accepted that synonymous mutations affect protein function. In a different approach, barcodes on cDNA constructs require a cumbersome second cloning step during library generation. We describe an approach that harnesses a highly sensitive duplex sequencing strategy to skip this barcode integration step and directly measure the target in label-free deep mutant pools. Furthermore, we employ the use of single-mutant standards with well-studied growth dynamics. By tracking the deviations from the expected growth trajectory of these mutant standards, we correct for dose-response variations that often result in low replicate agreement in these functional screens. Our workflow is compatible with any pool generation strategy, and the parameters that we obtain are robust enough to be useful in detailed models of cellular dynamics in synthetic biology studies. We developed a method for the functional assessment of deep mutant pools in a single pooled experiment. The workflow includes generating a label-free library of mutants at a desired target and sequencing the mutagenized pool before and during drug treatment using duplex sequencing. We demonstrate our pooled approach by focusing on target-driven resistance in Chronic Myeloid Leukemia (CML). To this end, we studied the 17 most clinically abundant BCRABL resistant mutations in imatinib refractory CML at low frequencies of 1 in 15,000. The sensitivity afforded by duplex sequencing enabled the accurate quantification of growth rates of these mutants under imatinib selection, with all 17 mutants closely matching their expected growth trajectory as predicted by individual dose-response studies. Moreover, we use an inference-based approach that uses hill-curve information from our mutant standards to detect slight variations in attempted dose between pooled replicates. Correcting for these slight variations in attempted dose leads to a >10-fold improvement in replicate agreement. Here, we demonstrate a fast, label-free workflow that achieves high levels of sensitivity and specificity in measuring growth dynamics of deep mutant pools. Importantly, our method does not rely on synonymous, potentially error-inducing labels. Given recent successes in deep learning that predict structure and function in proteins, our approach can be used to generate large training sets of mutation-drug phenotypes. Predicting the rich phenotypic data generated using existing sequence and structural data could dramatically reduce the experiments that are necessary to generate predictive information on drug resistance. Citation Format: Haider Inam, Scott Leighow, Justin Pritchard. Massively parallel functional assessment of label-free mutant pools is a universal approach to parametrize mechanistic models of drug resistance evolution [abstract]. In: Proceedings of the AACR Special Conference on the Evolutionary Dynamics in Carcinogenesis and Response to Therapy; 2022 Mar 14-17. Philadelphia (PA): AACR; Cancer Res 2022;82(10 Suppl):Abstract nr B014.