High allele frequency of somatic mutation

class: center, middle, inverse, title-slide

# High allele frequency of somatic mutation
### Jun Kang
### 2019 11 1

---

# Question 1

> Can we determine whether the variant is germline or somatic using allele frequency and tumor percentage in tumor only test?

---
# BRCA result

* BRCA1 mutation: Positive - p.Glu649Ter (c.1945G>T)  
* Variant allele frequency 63.84%  
* Tumor cell percentage: 70%  
* BRCA negative in blood sample

---
# Question 2

> Is it possible **64% allele frequency** of somatic mutation?

---
class: center
# Variant allele frequency in clinical tumor sample

`$$Allele \space frequency \approx Read \space count$$`
![:scale 70%](img/Allele_3.png)

---
class: center
# Somatic variant, Two copy, Tumor cellularity 50%

![](img/Allele_4.png)  
Variant allele frequency?

---
class: center
# Germline variant, Heterozigosity, Two copy, Tumor cellularity 25%

![](img/Allele_2.png)  
Variant allele frequency?

class: center
# Germline variant, Heterozigosity, Two copy, Tumor cellularity 50%

![](img/Allele_1.png)  
Variant allele frequency?

---
# Hypothesis

Germline, homo | Germline, hetero | Somatic
:-----: | :----: | :----------------------:
 100% |  50% |  half of tumor cellularity, `$\le 50\%$`

---
class: center
# Germline variant, Heterozigosity, One copy, LOH, Tumor cellularity 50%

![](img/Allele_5.png)  
Variant allele frequency?

---
class: center
# Germline variant, Heterozigosity, One copy, LOH, Tumor cellularity 25%

![](img/Allele_6.png)  
Variant allele frequency?

---
class: center
# Somatic variant, One copy, LOH, Tumor cellularity 50%

![](img/Allele_8.png)  
Variant allele frequency?

---
class: center
# Somatic variant, Amplification, Tumor cellularity 25%

![:scale 70%](img/Allele_9.png)  
Variant allele frequency?

---
# Variant allele frequency in clinical tumor sample

* Germline vs somatic  
* Tumor cell proportion  
* Loss of heterozygosity 
* Copy number

---
# Allele frequency in Somatic vs Germline in tumor only sample

* BRCA1 mutation: Positive - p.Glu649Ter (c.1945G>T)  
* Variant allele frequency 63.84%  
* Tumor cell percentage: 70%

---
# Allele frequency[1]

`$AF_{germline}$` `$=$` `${pV+1-p} \over {pC+2(1-p)}$`  
`$AF_{somatic}$` `$=$` `$pV \over pC+2(1-p)$`

* Given copy number (C)
* Variant allele count (V)
* Sample purity (p)
* Variant status (somatic or germline)

---
# Limited information

What we know | What we don't know
:-------------:|:-----------------:
Tumor cell percentage | Copy number
Variant allele frequency  | LOH

---
# Error

* Tumor cell percentage
* Allele frequency

`$$Allele \space frequency \not\approx Read \space count$$`

---
# BRCA mutation and LOH [2]

![:scale 80%](LOH.png)

---
# Allele frequency

* BRCA1 mutation: Positive - p.Glu649Ter (c.1945G>T)  
* Variant allele frequency 63.84%  
* Tumor cell percentage: 70% -> 80%  
* Somatic (supposed)  
* LOH (supposed)  
* One copy (supposed)

---
class: center
# Allele frequency

![:scale 70%](img/Allele_12.png)

`$AF_{somatic}$` `$=$` `$pV \over pC+2(1-p)$` `$= 0.67$`

---
# Allele frequency

* BRCA1 mutation: Positive - p.Glu649Ter (c.1945G>T)  
* Variant allele frequency 63.84%  
* Tumor cell percentage: 70% -> 60%  
* Germline (supposed)  
* LOH (supposed)  
* One copy (supposed)

---
class: center
# Allele frequency

![:scale 60%](img/Allele_13.png)

`$AF_{germline}$` `$=$` `${pV+1-p} \over {pC+2(1-p)}$` `$= 0.71$`

---
# Someone ask ...

>Does tumor only NGS test distinguish a variant somatic and germline?  
Answer 'No'

>Is it possible **64% allele frequency** of somatic BRCA variant?  
Answer 'Yes, It is when BRCA somatic variant accompanied with LOH'

---
class: center
# Germline variant, One copy, LOH, Tumor cellularity 100% (Bonus1)

Allele frequency 100% **hetero germline** variant

![:scale 60%](img/Allele_7.png)

---
class: center
# Somatic variant, Amplification, Hetero, Tumor cellularity 100% (Bonus2)

Allele frequency 90% **somatic** variant

![:scale 70%](img/Allele_11.png)

---
class: my-one-page-font
# References

[1] J. X. Sun, Y. He, E. Sanford, et al. "A computational approach
to distinguish somatic vs. germline origin of genomic alterations
from deep sequencing of cancer specimens without a matched
normal". En. In: _PLOS Computational Biology_ 14.2 (2018), p.
e1005965. ISSN: 1553-7358.

[2] M. M. de Jonge, D. Ruano, R. van Eijk, et al. "Validation and
Implementation of BRCA1/2 Variant Screening in Ovarian Tumor
Tissue". In: _The Journal of Molecular Diagnostics_ 20.5 (9.
2018), pp. 600-611. ISSN: 1525-1578.

.my-one-page-font {
  font-size: 18px;
}

.beginners{
visibility: hidden;
}

.intermediate{
visibility: hidden;
}

.advanced{
visibility: visible;
}
 {
    max-width: 100%;
}
</style>