Screening And GWAS Flashcards
What is new born screening ?
Screening is there to detect potentially fatal or disabling conditions as early as possible before signs or symptoms. It allows treatment to begin immediately reducing or eliminating the effect of the condition preventing devastating outcomes
Ins why genomic sequencing should be part of newborn screening
- Conditions could be screened for simultaneously
- Allow for interventions early in life for children with treatable or preventable diseases
- As soon as a genetic condition met the criteria for the newborn screening it could be added the next day
What is babyseq?
- The idea of babyseq is to explore the impact of incorporating genomic sequencing into newborn screening they looked at: 
- The medical impact on the individual and public health
- The behavioural impact on positions and patient behaviour
- The economic impact in terms of cost for the health care system

Barriers to implementation of poly genic risk scores into clinical use
- Confrontation of terminology
- inadequate description of applications
- Failure to define key elements of test strategy
- uncertainty as to evaluation approach
- lack of consensus on nature, quality and quantity of evidence
- concerns about the performance of applications in specific populations.
- imprecise descriptions of polygenic scores
- lack of evidence for their perceived value
- Concerns that they may exacerbate health inequalities
What is necessary for the successful clinical implementation of polygenic score based applications
That’s for implementation requires establishing evidence for the performance of each component of the test and how the test performed in a given healthcare pathway 
What is the importance of context specific evaluation in the use of polygenic scores
Context- specific evaluation is important because it defines the intended purpose,role and population for the test which is essential for developing evidence of clinical validity and utility
Give a definition for screening
Screening is a process used to detect early signs of a particular condition or disease in individuals who may not yet show symptoms. It aims to identify those at risk so that further diagnostic tests or interventions can be offered.
What are the key principles of screening
- Validity
- Reliability
- Acceptability
- Feasibility
- Early Detection
- Natural history
Describe why validity is a key principle of screening
Screening tests should accurately identified individuals with the condition and exclude those without it (High specificity)
Describe why reliability is a key principle of screening
Consistent results should be obtained when the same test is repeated
Describe why Acceptability is a key principle of screening
Screening should be acceptable to the target population
Describe why feasibility is a key principle of screening
The screening program should be practical, cost -effective, and feasible to implement
Describe why early detection is a key principle of screening
Screening aims to detect conditions early, when intervention can improve outcomes
Describe why natural history is a key principle of screening
UNderstanding the natural history of the condition helps determine the optimal timing for screening
What do you need to consider when putting together population genetic screening programmes
- Define the purpose of population genetic screening
- Explain how it differed from targeted or individualised testing
- Discuss the potential impact on public health
- What types of screening programs are you doing ? Universal screening, selective screening or cascade screening
- What condition are you screening for? Hereditary disorders or carrier screening or pharmacogenomics
- Implementation and logistics
- Ethical considerations ; address informed consent, privacy, and data sharing
What is stratified screening
Screening individuals based on risk assessment results. Based no stratification each group is given a different levels of screening.
Impact of population stratification vs screening the whole population.
- Fewer people are screened
- Reduction in harm associated with screening
- Reduced anxiety and inconvenience associated with testing
- Fewer false positives and unnecessary biopsies
- Less overdiagnosis
Common features of complex diseases
- Variation in multiple genes influence the trait, and have an additive, multiplicative or interactive effect
- Familial clustering with no clear pattern of inheritance
- Presence of a high-risk genetic variant doesn’t imply that disease will develop
- Often affects homeostasis gradually, resulting in onset later in life
- Often common in the general population.
How to find complex disease gene
- Linkage vs association
- Candidate gene vs Genome -wide
- Case control vs Quantitative trait
What’s the difference between linkage analysis vs association analysis
co-occurrence of a genetic variant with a disease trait, more frequently than can be readily explained by chance
Linkage analysis: looks at the cosegregation of alleles within affected FAMILY MEMBERS.
Association analysis: Difference in allele frequency between UNRELATED groups of affected and unaffected individuals
When choosing which test to do to find genetic factors of a common disorder what factors do you have to consider:
- Phenotype: Clinical definition
- Participant: Population based - will you use case-control or cohort
- Approach: Genome-wide or candidate gene
- Analysis: Lab analysis or Data analysis
- Verification: Replication or functional studies or Meta analysis
Define endophenotype also known as intermediate phenotype
A phenotype that you can quantify in disease cohort and control cohort eg for diabetes the enodo phenotype would be insulin levels
Or
It’s a measurable biological trait that reliably shows how a specific biological system works, is somewhat inherited, and is more directly linked to the underlying cause of a disease than just the general symptoms.
You need a large enough group to have statistical power to prove your hypothesis. What influences statistical power.
- Allele frequency
- Effect size
- Number of markers genotyped
How does allele frequency affect statistical power of a GWAS
-Higher Allele Frequency: Increases statistical power. When the allele of interest is more common in the population, it’s easier to detect an association with the trait or disease.
-Lower Allele Frequency: Decreases statistical power. Rare alleles are harder to detect, as fewer individuals carry them, making it more difficult to observe a significant association.
How does effect size affect the statistical power of GWAS
-Larger Effect Size: Increases statistical power. If the allele has a strong impact on the trait or disease, it’s easier to detect, even with a smaller sample size.
-Smaller Effect Size: Decreases statistical power. Alleles with subtle effects require larger sample sizes to detect significant associations.
How does number of markers genotyped affect statistical power
-More Markers: Generally decreases statistical power for individual markers due to the multiple testing burden. When more genetic markers are tested, the threshold for significance becomes stricter (to account for multiple comparisons), making it harder to detect associations unless the sample size is large enough to compensate.
-Fewer Markers: Increases statistical power per marker but may miss important associations not covered by the limited set of markers.
What is population stratification what affect does it have
Population stratification in GWAS refers to differences in allele frequencies between subgroups within a population that arise from ancestry rather than an association with the trait or disease being studied. These differences can lead to false-positive associations, where a genetic variant appears to be linked to the trait simply because it is more common in one subgroup than another, rather than because it is truly related to the trait. This confounding effect can distort the results of a GWAS,
What do you use to identify population stratification
A PCA plot
Define confounding factors
A confounding factor (or confounder) is a variable that influences both the independent variable (the factor being studied) and the dependent variable (the outcome of interest), creating a false association or masking a real one between them.
Why does failure to adjust for multiple testing appropriately lead to excessive false positives or overlook true positive signals.
- Excessive False Positives: When multiple tests are performed simultaneously (e.g., testing thousands of genetic variants for association with a trait), the likelihood of finding a statistically significant result by chance alone increases. If no adjustment is made, many of these “significant” results may actually be false positives, meaning they appear significant due to random variation rather than a true association.
- Overlooking True Positives: On the other hand, overly strict adjustments for multiple testing (e.g., using very conservative methods) can make it more difficult to detect real associations. The stringent criteria may reduce the power of the study, causing true positive signals to be missed because they don’t meet the adjusted significance threshold.
Most frequent phenotypes observed in infants
- Congenital and multiple congenital abnormalities
- Developmental delay
- Dysmorphism
Between WGS, WES & Gene panel which has ben shown to have a better yield (%) and changes in managment
WES has the greatest yield in infant diagnosis and leads to the most changes in management.
Why do a panel for infant diagnosis instead of a WGS if the yield is higher for WGS
The turn around time is shorter 100h vs 149hrs
Why is WGS more cost effective than the current standard of care methods
WGS is more comprehensive which means there is a higher probability of diagnosis and higher diagnostic yield which is more cost effective
List the pilot genome screening projects out there
- BeginNGS (USA)- WGS
- Generation study (England)- 490 genes associated with >200 conditions
- BabyScreen+ (Australia)- >500 treatable childhood -onset conditions
Screening vs diagnostic tests
- A proper screening program is that it is applied to a whole population
- Screening is a top-down approach offered to a large cohort
- Most screening tests do not results in a definite diagnosis with a higher rate of false positives and false negative outcomes vs diagnostic tests
- Screening tests need to be cheap and simple
- Diagnostic testing is aimed at an individual with a particular problem
Points to consider when asking when might screening be done?
- Can the subjects give properly informed consent?
- How easy is it to access the group to be screened?
- How relevant will the information be at that time to the person screened ?
- What are the practical and ethical implications of a positives results?
- What will the program cost, and do the benefits justify the cost?
Characterise population attributable risk
The proportion of the total disease risk in the population that is attributable to the factor in question.
Population attributable risk = proportion of the population that gave a variant (p) x the additional risk that variant gives (R) divided by overall risk in the population.
Common types of screening in healthcare
- Newborn screening
- Cancer screening
- Cardiovascular risk screening
- Genetic screening
- Eye screening checks in diabetics
Potential applications of polygenic risk score in cancer
- Risk prediction
- Tumour subtypes
-Contralateral breast cancer
-Familial or hereditary cancers - Screening
-risk stratified screening
-improving performance of current screening tests. - Guiding medical intervention
-information risk prediction of subtypes and selection of therapeutic interventions
-Health promotion
What needs to be considered about the implications of statisfied screening.
The size of the programme needs to be considered, for example under normal breast screening 7,4 million women are screened with stratified (Age) screening only 5.7 million women are screened. This means that there is a reduction in the number of cases screened which means there may be an impact on the number of potentially detectable cases.
Other the sample size what else has to be considered for stratified screening, what else is there.
- Which screening strategy i.e age or family history … or
- Which percentile or threshold should be used (example if its 70th percentile it means 70% of the population is within the risk category, So this means that if we stratify for age and end with 100 people and then only look at the 70th percentile we end up with 70 people of that 100)
Key questions when we ask if it is clinically useful to incorporate polygenic core information into healthcare for common disease.
1.how good are existing models to calculate a polygenic score
2. Does incorporation of this bio marker into risk prediction tools did value?
3. Does using this information lead to improvements in current prevention programme/pathway.
4. If this is better what are the organisational, ethical and social implications
Disagreements and challenges of using polygenic scores for screening
-Prediction accuracy: Disagreements on how well PGS predicts complex traits like disease risk.
-Research validity: Concerns about whether PGS research applies broadly, especially across different populations.
-Clinical usefulness: Uncertainty over whether PGS improves medical decision-making or patient care.
-Effective interventions: Questionable whether high-risk PGS results lead to actionable prevention or treatment.
-Evidence required: Lack of agreement on how much evidence is needed to prove PGS is reliable for clinical use.
-Purpose of PGS: Differing views on whether PGS should be used for broad screening or only in specific cases.
Why is transferability so poor for polygenic risk score
- True heterogeneity in allelic effect frequency
- Population specific variants
- Apparent heterogeneity in effect size
- Admixture effects
- Residual population stratification
6 Impact of rare variants missing from the PRS - Effect on gene-gene and gene environment interactions
Challenges associated with using genomic sequence data
- Interpretation is difficult and requires access to multiple data types
- Multidisciplinary expertise is often required
- Data storage is expensive
- Precarious funding reduced funds and increasing demand.
- Accurate diagnosis often relies on finding an unrelated genotype /phenotype match
6.knowledge is constantly evolving existing resources may be unreliable - Incidental findings of uncertain clinical relevance
- Lack of designated infrastructure
Accurate and early identification of disorders in newborns for diagnosing is difficult why?
- Atypical disease presentation
- Non-specific and overlapping phenotype (Facial features are easier to distinguish in adult faces not so much in newborn faces)
- Common phenotypic markers immeasurable in neonatal and newborn period
- Rapid phenotype and disease progression
- Obscured by comorbidities (sepsis, premature, birth trauma )
Clinical utility (measured by changes in clinical management and turn-around times) of an early diagnosis.
- Accurate prognoses and risk assessment
- Medication changes
- Initiation of palliative care
- Additional screening and monitoring
- Specialist consuls
- Supportive procedures and surgery
- Genetic counselling services
- Access to clinical trials for emerging therapies
Personal utility (subjective outcomes assessed across different domains) of an early diagnosis
- Affected utility -emotional state, acceptance and coping mechanisms, mental preparation, relief and responsibility
- Cognitive - information itself, knowledge condition and health status
- Behavioural - tactical utilisation, future planning, reproductive planning
- Social- social support, access to services
Characterise phenotype-first approach
- Guided by differential clinical diagnosis from a medical geneticist.
- This is a targeted testing and analysis which means decreased cost of testing /analysis but increased time to a result
- It’s reliant on strong phenotype-genotype relationships,
- limitations: genetic heterogeneity, limited phenotypic heterogeneity, requires doctors to have expertise about mild and non-classic phenotypes that most don’t have
Characterise genotype-first approach
- Examine genomic variation regardless of phenotype
- Analysis more comprehensive
- Useful for clinical heterogeneity
-atypical presentation
-varying severity
-conditions not fully penetrant
Facilitates new genotype-phenotype discovery
-new associations
-expand current phenotypic associations
-better understanding disease mechanism
Pros and cons Panel vs WGS
- Panels have a lower diagnostic yield that genome sequencing (research of 400 iII infants showed a yield of 27% for panel and 49%)
- Panels however have a lower turn around time
WES for infants, characterise
- Around 50% yield
- Only around 30% have been diagnosed by standard of care
- Around 20% of the entire cohort changed medical management
Pros and cons WES vs WGS
- They have similar diagnostic yields and turn around times
- WGS has better analytical performance, better nucleotide coverage which means 85x more rare variants called
WGS vs standard of care
Standard of care testing has a higher cost and longer hospitalisation and lower diagnostic yield
Characterise Rapid testing
- Faster results = earlier intervention (Fastest is 13.5 hours, decrease TAT by 35% )
- Involves automatic data transfer from electronic health records once test ordered, (Electronic health records are scanned for natural language/ terms that have known disease association ).
- Reduced library prep time, Faster sequencing , Pipeline used results in faster alignment and variant calling.
4.
New Born screening thats done in South Africa (not that important just read)
- Biotinidase deficiency
- Isovanic acidemia
- Propanic acidemia
- Galactosemia
- Maple syrup urine disease
- Glutamic acidemia type 1
- Vitamins b responsive methylmalonic acidemia
Controversy of newborn screening
- Interpretation genomic data constantly evolving
- Reporting on conditions later onset, psychological impact
- Associated costs - testing and counselling
4.access to therapies - most conditions are rare thus treatment expensive and inaccessible - Data handling and storage
What are beta values
SNP effect size estimates used to calculate PRS
DEFINE PRS
A score indicating an individuals genetic predisposition to a given phenotype based on a cumulative effect of many SNPs across the genome.
How do you calculate PRS
It’s the sum of risk allele an individual carries,weighted by the effect size of the allele.
Why estimate genetic risk
Personalised medicine
- Tailor prevention and treatment strategies based on genetic profile
- predict a persons prognosis or an individuals response to treatment
Research
-estimate genetic overlap between outcomes
-control for genetic effects
-investigate genetic architecture
Benefits of polygenic risk scores
- Easy to calculate
- Easy to store
- Stable across lifetime trajectory, inform a baseline risk
- Largely independent of traditional risk factors
- A single genotyping test to develop multiple PRS for different disease
Explain clumping
Clumping creates a list of approximately independent SNPs, identified in a manner that preferentially selects the SNPs most associated with the outcomes
Explain thresholding-
these independent SNPs are tested against a threshold to determine if they are included in the polygenic risk score or not
How does linkage disequilibrium affect the clumping algorithm ?
- We must account for linkage disequilibrium when polygenic scoring to avoid doubling counting highly correlated variants
- LD-based clumping removes variants highly correlated with the lead variant (SNP with lowest p value) within each locus
Limitation of the clumping method what is the way to fix this
Clumping may also remove non-overlapping information.
Joint Effect size- Rather than remove SNPs, estimate the joint effect of all the overlapping SNPs
Shrinkage - Keep all SNPs but effect size is shrunk toward 0 to make them more realistic and generalisable to all populations
The predictive utility of polygenic risk score depended on what ?
- Max variance explained by PRS depends on phenotypic heterogeneity ( example trait has high heritability, genetics plays a bigger role prs might be more useful and visa versa)
- Power/ GWAS sample size: polygenic scores explain more variance as GWAS sample size increase and genetic effects are estimated more accurately
What is the biological target of NIPT
Cell free fetal DNA which is short segments of DNA in maternal plasma primarily placental origin
detection rate and false positive rate of NIPT vs other screening techniques
- Detection rate of NIPT is Around a 60% higher than maternal age.
- Detection rate of NIPT is Around a 30% higher than biochemical screening
- Detection rate of NIPT is Around a 10% higher than nuchal translucency.
- False positive rate of other screening is 5% and NIPT is 0.1%
Biological challenges of cell free DNA NIPT
- False positives:
-unrecognised or vanishing twin (more DNA)
-placental mosaicism
- low level maternal mosaicism
- maternal genetic variation
-maternal malignancy - False negatives:
-low level of fetal DNA
-placental mosaicism
-maternal genetic variation - Failed results
-increased BMI
-low level of fetal DNA
-fetal aneuploidy
After NIPT gives Pos result, explain the diagnostic test that has to happen.
- Extract baby DNA - Amniocentesis (amniotic fluid taken 16 and 22 weeks ), chronic villus sampling (placenta taken 11 and 14 weeks)
- Perform QF-PCR or karyotype on the sample
Should we be implementing NIPT in South Africa
Implementing NIPT in South Africa could improve prenatal care by offering a safe, accurate, and non-invasive option for detecting genetic conditions, but factors like cost, accessibility, and healthcare infrastructure must be addressed for it to be practical and equitable.
Other possible uses of NIPT
- Other aneuploidies
- Gender
- Specific microdeletions e.g 22q11.2del
- Larger gains or loss of chromosomal material across the genome
- Single gene disorder
What are the cons of prenatal Chromosomal microarray compared to karyotype or QF-PCR
- Will not detect all genetic abnormalities
- Cannot detect balanced chromosome rearrangements such as reciprocal translocations and inversions (karyotype can)
- Cannnot detect low level mosaicism
- Anomalies may occur which cannot always be recognised or properly interpreted
- Cannot detect mutations within a single gene
When to Offer carrier screening
- Ethnic specific disorders
- For disorders cystic fibrosis and spinal muscular atrophy
- Women with family history
- Couples with consanguinity
