Sequencing
Use of Next-Generation Sequencing in the regulated domain of drug development
Next-Generation Sequencing is moving quickly from early research into the regulated domains of drug development, diagnostic development, and clinical decision-making. This article summarises some of the technical and regulatory challenges posed by these technologies and the efforts being made to address them.
N
ext-generation sequencing (NGS) has moved from the realm of research into those of clinical development, drug approval and clinical diagnostics, as the cost has decreased and the reliability of the underlying tech- nologies has increased. However, the process of translating raw reads into reliable genotypes is still subject to much variability. This variability pre- sents a challenge when using NGS in the regulated domain of the drug development process. Unlike more traditional biomedical assay tech- niques, or even other genomic technologies such as microarrays, interpreting NGS data depends on a long chain of data-processing steps after the raw data are generated. Each of these steps is the sub- ject of many competing algorithms, with more being developed and improved all the time. Furthermore, most algorithms have parameters designed to allow the algorithm to be ‘tuned’ to accommodate data generated under different experimental conditions. The result is that two independent analyses of the same underlying sequence data can lead to divergent conclusions. However, in a regulated environment such as a clinical trial or a treatment clinic, the goal is to have results that are robust and reproducible, and both analytically and clinically valid. New tech- nologies are being developed to help manage this problem and regulators are grappling with the nature of these algorithms in the context of their regulatory requirements.
Drug Discovery World Winter 2017/18 Algorithms
A typical NGS data processing pipeline includes the following steps:
1. The sequencing platform conducts image pro- cessing and the generation of raw reads (so-called ‘primary analysis’).
2. The next three steps shown in Table 1 (Read QC, Alignment or mapping and Variant calling) are often referred to as ‘secondary analysis’ of NGS data.
3. The last step (Variant annotation) is part of ‘ter- tiary analysis’ in which the detected variants are annotated and interpreted as to their likely biolog- ical or clinical impact.
Each step in the pipeline introduces its own opportunities for variability and generates quality metrics to help the analysts judge the usability of the pipeline’s outputs. Each individual nucleotide in a raw short read has an associated quality score that represents the likelihood that the base was identified correctly. Each aligned read also has a score that represents the likelihood that the read has been uniquely positioned within the reference genome. Each variant, and then each individual genotype, has a score that quantifies the uncertain- ty of the corresponding determination. Taken together, a sample analysis can follow any number of paths from a set of raw reads to a set of geno- types for the sample, with each path delivering dif- fering results.
In addition to these quality scores, algorithms 65
By Keith Nangle and Mike Furness
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72
