Massively Parallel Sequencing Flashcards
Content
- Principles of next generation sequencing/ massively parallel sequencing
- Why and how should MPS be used in forensic genetics?
- Advantages/ disadvantages with traditional methods
Why should we improve current STR workflows?
- Time from sample to result
- Cost
- Increase the information gained from analysis of a sample
Compare
- Time from sample to result
- Cost
- Increase the information gained from analysis of a sample
With: Rapid DNA, Direct PCR and MPS?
**Rapid DNA- what we’ve done
* Time from sample to result (yes)
* Cost (No)
* Increase the information gained from analysis of a sample (No)
**
Direct PCR- not doing DNA extraction or quantitation
* Time from sample to result (yes)
* Cost (yes- as no E or Q, but results may not be as good but does depend on samples)
* Increase the information gained from analysis of a sample (No- still DNA17 output)
Massively parallel sequencing (MPS)
* Time from sample to result (no- up to 48 hours)
* Cost (no- more expensive than what’s in current use)
* Increase the information gained from analysis of a sample (yes)
Advantages of MPS
*** Multiple tests run concurrently
* Improved marker discrimination
* Better relationship resolution
* Better analysis of degraded DNA
* Enhanced resolution of mixtures
* Identification of twins?
* Improved mitochondrial sequencing
* Metagenomics
* DNA intelligence tools **
Disadvantages of MPS
- Time
- Cost
- New interpretation challenges
- Nomenclature
Common MPS Platforms?
*** MiSeq- Illumina **
o Hospitals
o Medical research
o Used more commonly in forensics
* Ion PGM- Thermo Fisher
Explain how Ion PGM works
- DNA Template
- PCR
- Bind to bead
- Make many copies of template on bead
- Wash bead across porous chip, the bead will fall into holes
- Each hole individual PH meter
- Normal sequencing reaction done (template on bead), nucleotides incorporated and this gives H+ released when ntds bound
- Then pH meter can tell if H ion released or not
- Only 1 ntds at any one time
- A ntds washed across
- If A binds then H+ released and PH change
- If doesn’t bind then see nothing
- Remove excess Ntds
- Then wash with C ntds to see if H+ released
- Do multiple times to see what sequence is
Explain how MiSeq works
equencing by synthesis
2. Adding 1 base at a time
3. Flow cell (glass slide)
4. Glass slide has oligonucleotides on it
5. Synthetic DNA tail to bind to flow cell
6. Want to create multiple copies pf same template
7. Get clonal expansion on parts of flow cell
8. Now cluster all the same
9. With cluster, add sequencing primer, to sequence the template DNA
10. 6-8 bp tag added to know what samples are sequencing
11. Measures the fluorescence rather than PH like the Ion does
12. Add A, G, C, T but only one can bind, when it does get fluorescent signal
13. This signal shines and knows what ntd is incorporated
14. Different ntds binds after following washing
15. Etc.
How would one analyse multiple samples?
Specific index ligated during library preparation- different index for each sample e.g., 6-8 bp tag added to know what samples are sequencing in MiSeq
What is used to **run multiple tests concurrently **with MPS?
Autosomal STRs of ForenSeq illumine worksflow
Autosomal SNPs of ForenSeq illumine worksflow
o 94 identity SNPs
o Comparison of 47 SNPforID markers with GenPlex type
o Inheritance checked in confirmed paternity/ maternity
o Families
o Goof general heterozygous balance
Y-STRs of ForenSeq illumine worksflow
Condorance with ForenSeq illumine worksflow
Tell me about Improved marker discrimination with MPS
Why is looking at sequence variation such a benefit with MPS?
Able to see the variation between alleles which cant see with CE
Benefit of MPS with population frequencies
How many alleles do we need to type?
Yellow boxes show where further data was provided by NIST
As you can see that the more alleles are sequenced the greater the observation of alleles so many more to be found
STRSeq
LR distribution plots: half-siblings vs unrelated
- 10,000 simulations of LR shown in graph
- Density plot
- ESI= a DNA17 kit
- Want RHS of LR of 1 as are related
- LHS or LR=1 is not related
- LR=1 don’t know which hypothesis is correct
- As half-siblings there is a normal distribution spread across LR=1 as part is related/ unrelated
- This is what you get with normal CE testing
- Don’t want an overlap, but some will due to half-sibling sample. So you know that, the area in the overlap you could get unrelated pairs with the same ratio as related
- Not powerful enough to give full answers in all cases, hence use of MPS
CE vs MPS: half-siblings vs unrelated
- Same as before but with MPS shown in two new colours
o Moved further to L and R
o Still small overlap, but much bigger LR with MPS
LR distribution plots: half-siblings vs unrelated
- Same data set again
- But looking at MPS that is length based rather than sequenced based
- Looking at more markers with MPS and the ability to look at length based an sequenced based
Better relationship resolution with MPS
LR distribution plots: half-siblings vs unrelated
- Adding 94 identity SNPs + 27 STRs by sequences, pushes the two curves even further apart
- Now in billions for kinship relationships
- Hence more useful data using MPS due to sequence variation and able to use more markers which shows better relationship resolution
**Enhances resolution of mixtures **
**Enhances resolution of mixtures **
can also be done for example for….
* Two people major: minor mixture
* Three-person mixture
* Four-person mixture
DNA17- CE vs Sequencing
- Victim clearly in mixture
- Suspect A, B, D and E have varying levels of support to show they could be in the mixture
- Now look at same marks by sequence rather than size (blue)
- Suspect previously where didn’t have any data can exclude them, but shows what suspects are included
- So can see mixture is: Victim, A, D and E
**Better analysis of degraded DNA **
**Identification of twins? **
Do somemore reading around twins
**Implementation and results reporting **
