ath_RNASeq_reads.md

How many reads are enough for Arabidopsis RNA-Seq?

Table of Contents

• 1. Estimation
  • 1.1 Exons
  • 1.2 Transcripts
• 2. subSeq
• 3. References

1. Estimation

1.1 Exons

Estimate coverage from exons, and we assuming that the read length is 150bp

1. Retrieve exons from Ensembl database (Araport11). A total of 193,130 non-redundant exons with average length of 336.4bp.

   Coverage `Reads (million)`
   <chr>                <dbl>
 1 0.1X                0.0433
 2 0.5X                0.217 
 3 1X                  0.433 
 4 2X                  0.866 
 5 3X                  1.30  
 6 5X                  2.17  
 7 10X                 4.33  
 8 15X                 6.50  
 9 20X                 8.66  
10 30X                13.0   
11 40X                17.3   
12 50X                21.7   

2. Directly use the summary table from the TAIR database.

   Coverage `Reads (million)`
   <chr>                <dbl>
 1 0.1X                0.0414
 2 0.5X                0.207 
 3 1X                  0.414 
 4 2X                  0.829 
 5 3X                  1.24  
 6 5X                  2.07  
 7 10X                 4.14  
 8 15X                 6.22  
 9 20X                 8.29  
10 30X                12.4   
11 40X                16.6   
12 50X                20.7   

1.2 Transcripts

Retrieve transcripts from Ensembl database (Araport11), which including 3,746 lncRNA and 43,256 mRNA sequences.

   Coverage `Reads (million)`
   <chr>                <dbl>
 1 0.1X                0.0761
 2 0.5X                0.381 
 3 1X                  0.761 
 4 2X                  1.52  
 5 3X                  2.28  
 6 5X                  3.81  
 7 10X                 7.61  
 8 15X                11.4   
 9 20X                15.2   
10 30X                22.8   
11 40X                30.5   
12 50X                38.1 

2. subSeq

Randomly drop reads, then test the number of significantly changed genes, false discovery proportions (FDP), Spearman correlation and mean-squared error (MSE).

• flg22 vs. Mock

• flg22+SynCom33 vs. Mock

• flg22+SynCom35 vs. Mock

3. References

1. Robinson DG, Storey JD: subSeq: determining appropriate sequencing depth through efficient read subsampling. Bioinformatics. 2014;30(23):3424-6.