GTEx-sex-dimorphic-aging

the source code of GTEx sex dimorphic aging project

R version 4.0.4 (2021-02-15)

Required R packages:

R packages:

'impute=1.64.0';
'parallel=4.0.4'
'dplyr=1.2.2'
'limma=3.46.0'
'edgeR=3.32.1'
'sva=3.38.0'
'caret=6.0-93'
'caTools=1.18.2'
'zoo=1.8-10'
'forecast=8.16'

Functional enrichment analysis was performed by R packages:

'msigdbr=7.5.1'
'clusterProfiler=3.18.1'
'DOSE=3.16.0'.

Usage:

Rscript *.R or in RStudio

Data preprocessing:

Processing the quantification of gene expression (GE) and alternative splicing (AS) as judged by Paean

Gene expression:

TPM > 1 and protein coding genes

Alternative splicing:

(1) averaged total reads counts in spliced junctions (i.e., spliced in counts + spliced out counts) > 10
(2) non-constant PSI values across samples
(3) max(PSI) – min(PSI) > 0.05
(4) standard deviation > 0.01
(5) averaged PSI value in the range from 0.05 to 0.95
(6) TPM of the spliced genes >1

input: data matrix with genes/AS events in rows and sample name in columns
output: processed data matrix

PCA and pcSVR calculation:

Calculation of PCA and PCA-based Signal-to-Variation Ratio (pcSVR) to evaluate the contributions to global transcriptomic variations

input: GE/AS matrix and phenotype file (sex, age, etc.)
output: print pcSVR value and corresponding empirical p-value

Differential analysis:

sex-/age-differential analysis and sex-stratified differential analysis

input: GE/AS matrix and phenotype file (sex, age, etc.)
output: results of differential analysis in .txt file

columns:

coef_SEX2, coef_age_class2:Middle, coef_age_class3:Old, coef_SEX2:age_class2:Middle, coef_SEX2:age_class3:Old pval_age_class2:Middle, pval_age_class3:Old, pval_SEX2:age_class2:Middle, pval_SEX2:age_class3:Old fc_TPM_sex, fc_TPM_age_old, symbol

AD association analysis:

RandomForest classifier for predicting Alzheimer's Disease (AD) with feature selections procedure (RFE)

Sex-stratified model:

sBASEs in females predict female AD; sBASEs in males predicte male AD;

Control model:

Sex-stratified model by randomly selected AS events
Merge-sexes model: sBASEs in females/males predict AD

input: AS matrix with sBASEs in rows and phenotype file (e.g., AD, Control)
output: Performance of AD prediction in 100 iterations

SFs-AS events Network:

Construction the regulatory networks between splicing factors (SFs) and AS events

Step1: Spearman's correlation between SFs expression (TPM) and AS events (PSI) < 0.05
Step2: Significant regulation by SF Knockdown (shRNA-seq from ENCODE)
Step3: Existence of the SFs motif signals in the adjacent region (+/-300nt) of splice sites

Step1: Step1_correlation_SFs_and_AS_evnets.R:

input GE and AS data matrix; output spearman's rho and p-value;

Step2: Step2_diff_shRBP_AS_events_ENCODE.R:

input PSI matrix of shRNA-seq; output: differential AS events between control and RBP knockdown;

Step3: Step3_motif_identification_make_seqs_for_deepbind.R:

input list of AS events and genome annotation; output the subsequences (sliding window=40nt) in the range from upstream to downstream 300nt of each AS event to run deepbind;

Step4: integration of the results from the 3 steps above:

integration_networks_of_3_steps.R

Time series and Breakpoint analysis:

Identification of chronological genes and AS events during aging
Calculation breakpoints during aging
Identification of Aging-determiant Genes (ADGs) which can significantly alter the aging breakpoints

Time_series_analysis_for_chronological.R:

input GE/AS matrix; output chronological GE/AS during aging.

breakpoint_analysis_subsampled_genes_for_ADGs.R:

input hronological GE/AS during aging; output breakpoint and fitted TPM/PSI values.

ChenXi1201/Sex-dimorphic-aging