Work

Appendix Q: Recommendations for Developing Molecular Assays for Microbial Pathogen Detection Using Modern In Silico Approaches

Year: 2020

Type: article

Abstract: We describe the use of in silico approaches to improve the process of molecular assay development and reduce time and cost by utilizing available databases of whole genome pathogen sequences combined with modern bioinformatics and physical modeling tools. Well-characterized assays are needed for accurately detecting pathogens in environmental and patient samples and also for evaluation of the efficacy of a medical countermeasure that may be administered to patients. The polymerase chain reaction (PCR) remains the gold standard for pathogen detection due to the simplicity of its instrumentation, low cost of reagents, and outstanding limit of detection (LOD), sensitivity, and specificity. However, creation of such PCR assays often involves iterations of design, preliminary testing, and thorough validation with clinical isolates and testing in relevant matrices, which can be time consuming, costly, and result in suboptimal assays. Since formal validation (e.g., for Emergency Use Authorization [EUA] or Food and Drug Administration [FDA] licensure) of an infectious disease assay can be very expensive and can require extensive time of development, having a well-designed assay up front is a critical first step. Yet, many assays described in the literature utilized limited design capabilities and many initially promising assays fail the validation process, resulting in increased costs and timelines for successful product development. While the computational approaches outlined in this document by no means obviate the need for wet lab testing, they can reduce the amount of effort wasted on empirical optimization and iterative redesigns and also guide validation studies. The proposed computational approaches also result in higher performing assays with better sensitivity, specificity, and lower LOD and reduce the possibility of assay failure due to signature erosion. To provide clarity, an extensive glossary of defined terms is provided. Nucleic acid-based assays, such as real-time PCR, are the mainstay of clinical diagnostics and biosurveillance. A typical PCR assay design begins with computational (“in silico”) identification of a unique region (signature) that can support the binding of primer and probe sequences for target-specific amplification as a means of detecting the presence of the target organism. This step is followed by wet lab testing of the primers and probes using genomic deoxyribonucleic acid (DNA) or reverse transcribed ribonucleic acid (RNA) and performance-optimization of selected assays. In addition, extensive testing of the assay in the intended clinical matrix is required to evaluate assay parameters, such as LOD, sensitivity (probability of detection), and specificity (see glossary for definitions). The sensitivity and specificity of the assay are experimentally determined using a set of target (inclusivity) strains, near-neighbor (exclusivity) strains, and matrix-relevant (background) organisms. Assay performance also needs to be measured in assay-specific matrices (i.e., blood, stool, water, soil, etc.). Often, assays are computationally designed using a set of available genomic/gene sequences at that time and then experimentally validated for signature presence in all available samples of the target organism (inclusivity panel) and validated for signature absence in many other samples that do not contain the target (exclusivity panel and matrix panel). In an ideal scenario, a laboratory routinely engaged in assay development could complete this process within 6 to 12 months. Detection assays are typically designed using all sequences available at that time. Many of the biodefense assays were designed and tested at least a decade ago when available sequences were limited. Thanks to recent advances in modern sequencing technologies, there is a sharp increase in the availability of whole genome sequences (Figure 1). Hence, these older assays the to fail available of whole genome sequences for the assay design time available databases (e.g., typically contain a of a detection assays are to (i.e., and to target or detecting of the is limited for biodefense and as often and are other or and are In addition, such as the are (i.e., other are very and In detection assays to fail due to near-neighbor assays to fail due to target While the recent in sequencing combined with sequencing costs increased of genomic the for evaluation of not In this scenario, or older assays to of the target genomic is However, wet lab testing may not be due to on the availability of This is by such as the of that to biodefense pathogens for assay development and validation The on Detection for the such as and a of are to the for the detection assays. of the strains, and strains, are as a are limited to that are and for such extensive laboratory testing cost and time to the assay development Many whole genome sequences of these are available which the in silico evaluation of assays. set of assays to the sequencing with is in the of the assay to the target genome and to to the panel genome However, for all the assays there no assay (i.e., no and no assays were computationally to (e.g., assay in the panel) and (e.g., assay in the panel). Many of these assay to on the of these are other assays that fail the (e.g., assay or assay and are not diagnostics due to low specificity. However, the of the assay to the target in the panel and low in the the testing of all available is not cost described a of and for testing can be by in silico and in silico PCR testing of the target sequences for panel to a set the of in whole genome sequences of in and testing of molecular assays. are as up to by lower a A on and the of for testing assay development In addition, disease and due to and the of time in in the there a of 6 in for assay performance (Figure 6 to that there a in detection in that the assay the to this in or whole genome an assay could not be in a In the the availability of whole genome sequences within a time the identification of the in silico evaluation of the efficacy in detecting the of the for A in pathogen may the of signature and the development of or and detection assays, an assay the in a of of of of Hence, there is a for in silico and approaches to reduce the need for pathogen While in silico assay design approaches the need for testing pathogens in relevant in silico can be to the testing to isolates that are to assay failure to or of testing the to by the and While this document is not intended to or is intended to describe and for modern in silico assay design and and modern assay development are in the assay design the is on laboratory wet lab The of the modern process are extensive use of in silico of whole genome sequences to guide and the of the and environmental panel wet lab testing, on and cost and time assay the modern is and and using assay design to evaluate assay and assays to wet lab testing This the of iterations to the The and the of the modern assay development assay target an process that is by the and of assay Often, assay are selected on of pathogens or or The resulting assays are then computationally or for and using The resulting assays are also often computationally all sequences by using to this is not many that are not relevant to sequences that be in relevant matrices (e.g., or In other primers all sequences is to In a modern an of all available for the organism of to and then and environmental (e.g., environmental panel of design is a critical in the development of assays that to other of the process, such as instrumentation, and However, primer design a to improve performance (i.e., sensitivity, specificity, and as as reduce the development time and a primer design a design that for intended were not to be for primer a result of the of such a assay extensive with of and testing and the resulting assays are often and to failure modern are for step in the assay development are physical and modern design and computational to of the with the primer design The are and not an of tools. other with an can also be for the to the of PCR is the availability of modern are a that can be to improve the design of the PCR assay to (i.e., the of that are by a PCR and also guide the testing of a PCR assay by or (e.g., due to at primer and probe a PCR is to first of sequences that the and (see glossary for definitions). In to the such as for and the of there are a of pathogen genome such as the and and We also using primer design that such databases as an of design, such as and to the of for genome databases the of and to the at in a genome the is is by a of in strains, sequencing and can for which a of the genome is available the product of sequencing or as as sequences that contain or by the of and are in clinical and or that can to a or the is to sequences in the determined sequences can reduce the of in a set of target Use of sequences in the can assay design to promising and design design by the signature (e.g., due to the of sequences are on that are Use of sequences that contain or sequences can result in assays that sequences that do not in The ideal when the the of pathogens with (e.g., and A The availability of sequencing such often such a not is on the assay to all available sequences a the (see the A and and such utilizing is of for PCR there are (e.g., and a to with infectious or that these utilizing the result in many to be in sequencing that many of are not for primer is to as as and in these However, as assay design to the of sequences with the of assays, sequences assays to the that on the of the genome that are While this can be a when with for which is better by available sequences available whole genome is a the available sequences are a region or a region that is in is to and by and and sequences also for many design they a defined and they for on a sequencing can the with is to up front to the in the for all of the also due to genomic that is to that of on computational and for signature describe that are of is that be a the can then be for of design in testing for with an such as or using a combined of and In there is an of sequencing for a (e.g., a or a an is to in the sequences or that contain that of The of and in to (Figure The low cost of sequences using sequencing to increased of of complete can be by these with sequences such as or with there is a time in the of available that are complete (Figure the with (e.g., there is an increase in of time. with the increase in the of databases be a on the computational to such the of and complete in and the that are complete as a of were by the that as an of all on the are on the genome The were on the of for is to and environmental computational the with near-neighbor sequences (i.e., that are to in the other that may be in the matrix and can be the that the of that the that can be in a clinical matrix or other (e.g., in an environmental etc.). the and is not an and is to sequences as as complete A is to primers for with all as the or using a such as or primer to all and However, is not to use such databases the design the and databases contain many sequences that no possibility of in the matrix and such sequences provide design that are not and could result in a suboptimal the first step in design is to a region of the pathogen genome that is of a A (e.g., or is the to such However, do not terms of and with of sequences or with with modern a can be In addition, pathogen and sequences in of and combined with the low of (i.e., for for is to the that are required to the to an to of or (e.g., and of the design are for the and provide a a design region is with a do using an that is to the design region of not for the A for the design is to using (i.e., of on the the is described in and are computationally for databases and sequences and can be to databases of and design region a pathogen the of the target (e.g., clinical isolates of a and a of to near-neighbor or to that could the use of to and genome databases to of signature design such is described in the literature by and to the development of an assay for the a is also available in the approaches or genome to unique that can be assay In all these the first step is to the and environmental the of primer is critical for assay design design as available many to the are the standard in at standard and standard and In performing such on the to is is the at which of the target is by the and is However, not the amount of at the for primers or at the for A is that the to design primers is to This is for the are the binding to and at the and the not the and target can be using such as or on is to use that on the for the amount at the The a of that use to a amount to a amount of the amount is using a In to and is to of primers to that they do not the of the This can be with such as and are also a of approaches for the is to primer a of to the primer is to using and is outstanding for such However, for primer design, is not the that the of the to a primer is the that for primer A better is to use (i.e., or the amount the for and as as the of and of that of genome databases are and A of these is to not also to evaluate the are by a polymerase (i.e., at the of the and of primers are in and within (e.g., that all are (e.g., the are not all at detecting all (e.g., such as do not very for detecting a PCR reaction the use of a of such as many is for testing the of primers for of a amplification (i.e., a with an and in the presence of target genomic and performing However, such all amplification (i.e., the and and are for the use of detection is not for assays. that the is the target of sequencing (e.g., sequencing assays, the use of an probe (e.g., molecular or an of specificity in that that to the probe are such are the target of the detection with the detection can provide that an assay is performing The design for probes and not be probes such as and and to intended that probe sequences can be to probes that contain probes can be for and that do not a signature region However, a of such and probes is that they can fail to to of the target that contain such probes to signature erosion. to use then a that the of is an A critical of primer design is the of the using of there is no for of primer there are many that such as for and amount target primer and a of is of an to all of these terms and a result to a as a that to In addition, there is no in the as to the of terms this is to use that the and the for (e.g., a of the by the to the and also to be the modeling is and is not In the when and validation available as described in the terms and can be by for the also support the to evaluate the of and In provide an of a design that the of the modern described in this modern is the use of databases for and are in the target as as the of to and specificity. is also physical modeling (e.g., such as or to for and and for the to the amount is also the critical for the using a a of design that be in a modern primer design these require the described in the the for a is in the In addition, an iterative process is for performing validation to assays. for Detection by of for PCR primer design of for PCR primer design involves performing assays in the reaction the of the of reagents, and also the amount of can be as as the target is PCR in the presence of an (e.g., or to pathogens are in the The of is to of primers and probes that are a set of reaction means that the primer with do not to do not and do not the matrix the primers to at is critical to that amplification of or not the reaction and all the or to all of the amplification can be using the described to design primers that to (i.e., of the result in that do not that can polymerase and primers that do not testing of to that and not a in the reaction is to the of is to computationally However, the in the of computationally to use a to all for all that can (see PCR is a with many (e.g., genomic the of and the set of primers can result in an with set of primers in the reaction to the and databases or In addition, as the of the reaction there is a in the of a reaction with primer design (i.e., for of the the of to is iterative empirical can a of the and samples a very limited amount of the design resulting in a suboptimal design and often complete an iterative computational using that could of of the of the with be computationally to (e.g., or for all to that are a modern is needed to the and the The a with a to the design and the of the of We using such optimization to reduce the of iterations required to a modeling is and there are many that are in the design However, using such design result in many design and iterations and assay is the of and an often in the development and validation of in silico PCR assay design that increase and with a for can also be to in silico of the in PCR assay design is the of an assay to or While the needed to a PCR assay on the of the computational relevant can primer and probe sequences and sequences and and polymerase and (i.e., and these can the and to the that is the with the However, is often to for many of these in polymerase and the may be that are not can be or such as are often measured in that (e.g., be the of required for PCR with or In the absence of be determined by to of PCR that are not to (e.g., reaction and also such as and for primer design, of for primers with to improve and of sequences that be or the physical are not in the possibility of to PCR assay on a of physical process (e.g., and assay of they are PCR an with in to and assay performance to to wet lab for is the of a PCR are the and the PCR can be determined the PCR (e.g., by the a of target the is a for the amplification of the that result in are to be to primers that result in higher that due to or of primers due to amplification all result in higher with a an limit of detection the of PCR is limited is is to no of PCR measured is the of the PCR (i.e., in the first of the PCR PCR a PCR and due to performance of the probes or of the primers due to amplification However, there is no to the and a there is no to the and to is a in a PCR the and the are the of the primers and the to as the and are for the primer design there is a need for the to that all the in the of a real-time PCR A in this is modeling the of the PCR reaction using the and in the which modeling of the PCR for PCR for PCR In to the and needed to and in silico PCR of are needed to and can evaluate the computational in While computational of PCR within the of and PCR, many do not this a set is needed to The set could be to evaluate The PCR set need to many target presence of and primer and probe of the sequencing to the of the and primer and of the PCR reaction at of PCR (e.g., real-time of the with of primer and this set available the PCR and be to the to improve and in silico to the of there is a need for an to the performance of computational approaches for in silico PCR using could provide a of by and the development of in silico In to in silico be to experimentally real-time PCR assays, that are with in the and reverse and of and the the PCR and in be the validation the PCR be be for evaluation of PCR by The be to evaluate the performance of in silico PCR (i.e., and To there not an effort to provide to the needed and validation PCR an to these or of to in the of this effort the and a on the of an assay designed in the step is to the performance of the design in wet lab testing using and the genomic in genome databases resulting sequencing many assays require the of primers probes for assays require design on the intended use of the genomic need to be and the is genome The presence of in the primer or probe binding can the PCR (i.e., the for a target and the of target need to be determined for target to sequencing the to the primers and probe are an to the organism assays are designed and and performing primer are or for or optimization of or assay a organism or organism or or a with the assay target or assay target is as at this of the assay development primer are is needed to assay can be in a matrix such as validation in the intended sensitivity the assay LOD, or the of organism that is (e.g., when of are specificity involves and (exclusivity) testing, all the available or of the organism are tested at The selected for testing within the assay target acid could also be to the assay and using whole organism. testing other within the target acid the intended matrix (e.g., whole and that were the that that testing that While there are for assay validation that can be on the the described and in a of for PCR assay validation for clinical Assay validation a of the (i.e., the to process all such as testing, and to the assay performance in the intended and the is validated for that and assays require sensitivity and specificity, and testing with and testing the testing of sensitivity in the intended matrix is to a preliminary LOD with a of LOD, for at least samples Assay the of an assay when testing of a the specificity testing to an of (e.g., whole blood, on assay assay and LOD in the intended matrix to a matrix can that testing as many as (e.g., time PCR etc.). on and testing samples the LOD and and the for the result is to the In a gold standard is a to the assay can be In such a is not a clinical or can be samples at LOD, at a and at a relevant and samples are tested testing for assays is to assays the samples are at a a low and at the assay LOD In to the assays require the of the that the of the standard can be accurately the lower limit of is by testing samples in at the within the and the at least of the are with a of The limit of can be determined While within this can be can also or that are this an assay is designed and is for the to a of a pathogen that a at a that the assay to This process is To there not for is to design primers to be a needed for a In the that a or at a the primers as as the not at the of In addition, there is a need to that can computationally validated assays to the is to be by the assay (i.e., the assay is to an for the with or the assay is to fail for the (i.e., the assay is to an for the with low better be for to databases for and to a or a is for assay This for real-time of assays to in of the assay The could then validation on assays that a of failure using limited on development, and of such an assay is by the In silico can be for assays to a of clinical In silico approaches such as outlined provide for development of molecular assays in provide unique validation in wet testing is in silico can that can be validated in testing to the of (i.e., of that can assay which on assay specificity or (e.g., the to which can be in silico are the design, development, and validation of an assay for clinical use on that the of the intended of the assay (e.g., acid-based and of these also the of a for for to The is a on the of a can for a medical and is intended to as an on (i.e., a molecular assay or that is within the of this can be to in the process they that they to be by the on medical on the process is of the and the may the use of medical or of medical for may an the and first that the on of by the of of or of of a an or a a and or a disease or that may be to such can be in the Use Authorization of and To for and with medical countermeasure to when A and the and efficacy of the its intended use a or and the or The process to and in the development of of and other needed for an in of an and also to of an a that a can to an there is a on to a for in diagnostics to can be at the at the of the may the of a medical product to be in an to or or or by when there are no and available that are to an for a medical the and of the product the and on the of the at the time of the The the availability and use of needed to for and to an is by the then on that no in and can no be The for medical (i.e., assay is to of and to to a medical The is on of medical This is intended as an to medical development and to the in which they can for on (e.g., clinical or to that may to the for a such as the on such as validation to support intended use and in the The to the in to that all that were in at that time of or the of to provide of and The which a is the that a medical to a in to of the the are defined as the are to a set of that are to all of for which by are to provide of the and of the for which there is to to provide such for which by are and for which there is to to provide of the and of the typically require is the process by which a (i.e., is of the A to in the a or intended for for which a is not to a to as a the is the of the and not the of for of the of In a the or not the the for The its on the is (i.e., as and as

Source: Journal of AOAC International

DOI: 10.1093/jaoacint/qsaa045