바이오 대표

[Bioinformatics 논문] TIN score 이용하여 RNA degradation 계산 “Measure transcript integrity using RNA-seq data” 본문

논문

[Bioinformatics 논문] TIN score 이용하여 RNA degradation 계산 “Measure transcript integrity using RNA-seq data”

바이오 대표 2023. 2. 3. 04:59

“Measure transcript integrity using RNA-seq data”

Feb.2016

 

Abstract

RNAs 데이터는 보통 archived (저장되어있던) clinical tissue 에서 추출하는데 꽤 degraded 한모습을 보인다. RNA degradation은 RNA-seq read coverage를 왜곡하기도 하고, whole-genome gene expression profiling 에 심각한 영향을 끼치기도 한다.

 

Result

TIN (transcript integrity number) 을 이용해서 RNA degradtion 을 measure 할 수 있다. Calibrating gene expression counts with TIN scores 를 통해서 by false positive 를 줄이고, 유의미한 pathway를 복구함으로 효과적으로 RNA degradation effects 를 neutralize 할 수있었다.

 

예전 방법

  • gell electrophoresis (ribosome RNA band 육안검사)
  • capillary electrophoresis (모세관 전기영동법)

Conclusion

🙂 TIN 좋다.

 


 

Background

표본 collection / storage condition 에 따라 RNA sample 의 degradation 정도가 결정된다. 이렇게 손상된 RNA 는 in vivo gene expression 을 정확히 measure 하는데 어려움이 있다. 유전자의 발현양을 측정하기 위해, hybridization-based microarray platform 을 이용했을 때는 a few short, discrete probes를 사용하기에 RNA degradation이 거의 없지만 sequencing-based RNAseq 을 사용하면서 꽤 RNA degradation을 보인다. [5 - showed up to 56%]

 

RNA Integrity Number (RIN) 측정에 사용되는 요소들: “total RNA ratio”, “28S-region height”, “28S area ratio” and the “18S:28S ratio”. 해당 측정은 rRNA 가 메인이기때문에 이것만 가지고 mRNA를 사용할수는 없다. 또한 RNA decay rate 자체가 functional groups / transcripts 마다 다르다. 또한 해당 수치는 overall measurement 이기 때문에 Differential expression을 측정하는데에는 충분치 않다.

그래서 RNAseq 데이터를 이용하여 read depth를 이용하는 transcript integrity number (TIN) 계산 방법 개발했다. As a result, TIN metric는 mRNA integrity 를 measure 할 수 있고 high concordance with RNA fragment size.

또한 해당 TIN 을 이용하여 gene expression 을 adjust 할 수 있고, False positive 를 줄임으로써, DE analysis 를 improve 할 수 있다. TIN = “percentage of transcripts that have uniform read coverage across the genome” 로 정의할 수 있다. 

 

 

Results

3’ biase - RNA degradataion by 5’ exonuclease & oligo (dT) selection will only isolate most 3’ portion

 

Discussion

RIN - pre-sequencing sample screening

TIN - sequenced RNAseq data as input

 

Methods

  1. Datasets
  2. Determine the RNA integrity number (RIN)
  3. RNA samples analyzed by Agilent Bioanalyzer 2100
  4. Transcript integrity number TIN compute
  5. using relative coverage (P_i) and Shannon’s entropy
  6. Calculating library RNA fragment size
  • TIN = -10 log10 (standard deviation of the fragment length distribution). A TIN value of 10 or higher is considered good quality, while a value below 7~8 is considered poor quality.

실습

https://bioinformaticsworkbook.org/dataAnalysis/RNA-Seq/Identify_Strandedness_Information.html#gsc.tab=0

https://liulab-dfci.github.io/RIMA/Preprocessing.html