Scribd
Upload
2 views

Supplementary Table 2 Assay Design and Test Optimization Considerations

This document provides a structured approach for designing a new assay, starting with defining assay requirements and progressing through initial assay design, performance assessment, and validation. It emphasizes the importance of tracking specimen identity, confirming variants, and iterating on design based on test results. Additionally, it outlines the need for acceptance and rejection criteria to ensure the assay meets specified requirements.

Uploaded by

aa
Copyright
© © All Rights Reserved
Available Formats
Download as XLSX, PDF, TXT or read online on Scribd
2 views

Supplementary Table 2 Assay Design and Test Optimization Considerations

This document provides a structured approach for designing a new assay, starting with defining assay requirements and progressing through initial assay design, performance assessment, and validation. It emphasizes the importance of tracking specimen identity, confirming variants, and iterating on design based on test results. Additionally, it outlines the need for acceptance and rejection criteria to ensure the assay meets specified requirements.

Uploaded by

aa
Copyright
© © All Rights Reserved
Available Formats
Download as XLSX, PDF, TXT or read online on Scribd
You are on page 1/ 14

How to use this workboo

1. Start on tab A, Assay Requirements

The most important aspect of designing a new test is the setting of requirements. This will inform all further work a
fill in the genes or variants that are to be included in the test. Note if each gene/variant is absolutely required to en
Continue with other requirements around turn around time, sample material to be validated (e.g., saliva, blood), co
suggestions for requirements, but there are likely many other aspects that are unique to each test or laboratory. Re
be carefully discussed, documented, and signed off by all stakeholders.

2. Continue to tab B, Initial Assay Design


This phase translates the requirements to an initial assay design. The target calculator estimates the size of the gen
combination, how many samples can be sequenced per run and how many samples can be comfortably processed p
each sequencing instruments, with varying data outputs, costs and run times. If the assay design involves purchasin
these factors as well as requirements for future scaling. Since these offerings change regularly, first contact instrum
possibilities and limitations. Another important consideration is the assay methodology; i.e., amplicon vs. capture b
interrogated at the same time using a capture based assay. Amplicon technology is very useful for technically challe
flexibility offered by PCR optimization possibilites is required (for instance, to address GC rich regions). There are lim
but note that various commercial offerings extend these limits and improve workflow, for instance by using fluidics
primer/primer interactions. Capture based technologies offer more flexibility in terms of parallelizing multiple tests
but have limitations on optimization possibilities. As a rule of thumb, if more than 15 genes are planned for the test
capture based assay may be the best choice. Choose both the sequencing instrument and the methology after care

Specimen identity should be tracked in order to detect sample mix-ups. Confirmatory testing of reported variants is
the identity check to all samples, more extensive QC such as testing a pre-determined set of common variants (finge

3. Return to tab A, Assay Requirements


Ensure that the completed Initial Assay Design will theorically fulfill all "must have" requirements. If not, reassess a

4. Continue to Tab C, Assay Details

On this tab, list the design and samples you will use to assess the performance of the assay in the first test run. This
to define the path of optimization reach fulfillment. Be conservative when choosing the number of samples to run, a
order to be able to pool the data from identical samples to simulate a decrease in the sample number. Choose a lim
reserve the majority of samples for validation. Prepare the draft SOP before processing any samples, and while perf
optimization. It is important to process the samples in a manner as close to the envisioned final assay as possible - i
non-reproducible manner. Once the run is complete, it is recommended to analyze the data in multiple ways (see b
5. Iterate as necessary

After performing the first test run, carefully assess the metrics against the Assay Requirements on tab A. If any of th
them. Alternatively, additional research may be needed to decide if a given requirement is truly necessary. Assay a
entering into the validation phase. Once the assay satisfies all necessary requirements, the information on Tab C an

6. Define acceptance and rejection criteria (pass/fail metrics)


Acceptance and rejection criteria for the results generated by the wet bench analytical process (MOL.36010). Criteri
optimization and utilized during validation. List the criteria on Tab C, Assay Details.

© 2019 College of American Pathologists (CAP). All rights reserved.


Test Requirem
Test Requirements
Before you begin, carefully define and document the test requirements
Gene region 1 to be covered
Gene region 2 to be covered (etc.)
Necessary sample throughput per week/month?
Maximum possible cost of goods (all inclusive)?
How deeply does each position need to be covered for accurate variant calling (if known - otherwise address during test optim
Sample material 1 (e.g., saliva)
Sample material 2 (e.g., EDTA whole blood) - etc.
Are there known problem regions that require additional testing?
Required/expected TAT?
Combine different tests (existing or planned) within a sequencing run?

Add other requirements as needed, e.g.,


Specific market expectations that need to be met
Restrictions due to existing instrumentation
Expectations for scaling
Limitations due to informatics infrastructure
must have nice to have reasons for decision
Initial Ass

1. Sequencing instrument table: obtain informati


Vendor Instrument Kit Output
Illumina MiniSeq
Illumina MiSeq kit1

Illumina MiSeq kit2


Illumina NextSeq 500 kit1
Illumina NextSeq 500 kit2
Illumina HiSeqX
Illumina NovaSeq
Ion Torrent S5/S5 XL
Ion Torrent Proton
Ion Torrent PGM
Pacific Biosciences RS II
etc.

2. Target calculator: Determ


Number of Size of coding
exons region
Gene 1
Gene 2
Gene 3
etc.
total number total size of target
of exons

total number
of exons x 2
Buffer zone outside each exon (50 bp suggested) x 50 bp

total size of
target +
Total target size (bp) buffer zone

3. Estimate how many

Output of sequencing run (bases)


Desired average coverage (X)
Theoretical number of samples/run Output/average coverage
Theoretical number of
Expected number of samples/run samples/run divided by 2
Expected number of
samples/run multiplied by
Expected monthly throughput runs/month

4. Define confirmatory/sam

Are variants confirmed prior to reporting? yes/no


Assay to be used

Is sample identity tracked for all samples? yes/no


Assay to be used

5. Make initial cost

Run cost divided by expected number of samples/run


Cost of extraction
Library preparation costs
Additional consumables
Confirmatory and/or supplemental assay costs
Labor estimate
Overhead estimate (incl. instrument depreciation)

Total cost per sample Total of above

Total cost with 10% rework rate Total cost multiplied by 1.1

6. Design review: compa

Copy in all requirements from tab A and check if met

Cost reduction options include: different


sequencer/kit, different library preparation kit or
method, eliminating nice-to-have target regions etc.
If the cost is higher than the requirement but
reasonably close to meeting expectations, bear in
mind that several conservative assumptions were
made. It is likely worth proceeding with the test run
to assess actual vs. expected metrics
Initial Assay Design

ment table: obtain information from vendors and literature to assess options
Time Cost/instrumentCost/service contract Cost/kit

2. Target calculator: Determine your target region size

3. Estimate how many samples can fit on a run


4. Define confirmatory/sample identity testing approach

5. Make initial cost estimate per sample

6. Design review: compare against requirements


to assess options
Limitations

Vendor only allows for use with whole genome sequencing


A very large number of samples must be pooled to be cost effective

Notes

Buffer zone allows for flexibility in bait or primer design

Choose an appropriate instrument and kit from the sequencing


instrument table
Choose the desired average coverage (100X is a good starting point for
germline tests, but expect to refine after performing the test run)
A good rule of thumb is to decrease by 50% from the theoretical number
of samples to accommodate uneven coverage, pooling, and suboptimal
sequencing runs

Start with a conservative rework rate as specimens often need additional


processing
Assay Details

1: General Information
Collect details on primer or capture probe sequences here:
Probe/primer sequences, transcripts, source, location.

2. Supplemental Assay details (if known - otherwise add information after first test run(s))
List any known or expected difficult-to-sequence regions and collect details on supplemental assay(s) here.
Examples include high or low GC content, highly homologous regions, copy number variants, and repeat expansions

3. Sanger confirmation and/or fill-in (areas that may need fill-in may be determined from the first test
List areas with known or expected NGS failure that will be covered by Sanger fill-in sequencing, if any, and collect de
If variant confirmation is to be performed, define the specific policy and list pre-designed Sanger primers here.

4. Samples for first test run


List samples for first test run here. Ensure that a second, independent, sample set is available for validation.
Include both known positive and negative samples, and test all specimen types.
For large panels, consider including a sample with an available public data set, such as NIST. Reserve most of these

5. Define acceptance and rejection criteria


List pass/fail metrics to be used during the validation and in production here.
for the validation.

You might also like