Backdown

2025-02-04

Remarks : use Biostrings or seqinr packages to lighten code

library(tidyverse)
library(forcats)  # fct_reorder
library(tibble)
library(data.table)

source("R/functions.R")
source("R/extradata.R")
source("R/functionsVisualisation.R")

Nucleotides Sequence creation

Sequence itself

S <- genRnaSeq(n = 40, complementary = TRUE)

Codons splitting and shift

## Shift = 1:

## List of 2
##  $ Original     : chr [1:13] "CCC" "UUA" "CCA" "UUA" ...
##  $ Complementary: chr [1:13] "GGG" "AAU" "GGU" "AAU" ...

## 
## Shift = 2:

## List of 2
##  $ Original     : chr [1:13] "CCU" "UAC" "CAU" "UAC" ...
##  $ Complementary: chr [1:13] "GGA" "AUG" "GUA" "AUG" ...

## 
## Shift = 3:

## List of 2
##  $ Original     : chr [1:12] "CUU" "ACC" "AUU" "ACU" ...
##  $ Complementary: chr [1:12] "GAA" "UGG" "UAA" "UGA" ...

Data creation and curation

Here’s a glimpse of the data we have for each length

## # A tibble: 9 × 6
## # Groups:   Length [9]
##   Length AminoAcid   Type  TotalCount Proportion MeanTm
##    <dbl> <chr>       <chr>      <dbl>      <dbl>  <dbl>
## 1    100 Cysteine    Comp…          6     0.0192   9   
## 2    150 Valine      Comp…         30     0.0642   9   
## 3    200 Tryptophan  Orig…         12     0.0191  10   
## 4    250 Leucine     Orig…         68     0.0864   8.33
## 5    300 Isoleucine  Comp…         57     0.0593   6.67
## 6    350 Lysine      Orig…         41     0.0371   7   
## 7    400 Arginine    Orig…        124     0.0983  10.3 
## 8    450 Tyrosine    Comp…         57     0.0397   7   
## 9    500 Aspartic A… Orig…         48     0.0303   9

Visualisation

Amino Acid Proportion Distribution

Hypothesis 1 : It appears that there is a clear correlation between the number of codons encoding each amino acid and their observed proportions. To confirm this, we assess the monotonic positive relationship between these two variables using a Spearman correlation index.

We display a heatmap of absolute differences in the original and complementary strands frequencies

Hypothesis 2 : The ‘random’ distribution of AA tends to be very close between the original and complementary strand as the size of the strand and the number of repetitions increases

Amino Acid Melting Temperature Distribution

Hypothesis 3 : The data suggests a positive correlation between melting temperature and amino acid proportion. However, no definitive conclusion can be drawn, as the melting temperature calculation process is highly questionable and requires further scrutiny.