Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/wickdchromosome/leastsq_afc
Use non-linear optimization to estimate allelic Fold Change
https://github.com/wickdchromosome/leastsq_afc
Last synced: 26 days ago
JSON representation
Use non-linear optimization to estimate allelic Fold Change
- Host: GitHub
- URL: https://github.com/wickdchromosome/leastsq_afc
- Owner: wickdChromosome
- Created: 2019-07-18T23:25:22.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2019-07-26T15:42:59.000Z (over 5 years ago)
- Last Synced: 2024-03-12T15:29:09.305Z (8 months ago)
- Language: Python
- Size: 7.33 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Least-squares aFC
## Overview
This script uses a non-least squares fit to calculate allelic fold change.
It calculates the linear approximation for the aFCs for comparison.
The models used are based on the paper https://genome.cshlp.org/content/27/11/1872.short .
Both the vcf file and covariates file are in the standard format for FastQTL: http://fastqtl.sourceforge.net/
This project depends on: Python3, pysam, pandas, datetime, numpy, argparse, gzip, lmfit, sklearn
This project was tested on Debian 10.
## Usage
```
spheus --flag1 --flag2
```
### Use flags
#### Required
```
--vcf The VCF file with phasing data for each variant ID
--expr The expressions file with expressions data for each individual for each gene
--eqtl The eqtl file tie
--isnorm -n Are the expressions normalized already? 1 for yes, 0 for no
--islog -n Are the expressions log transformed using log2 already? 1 for yes, 0 for no
--output -o
```
#### Optional
```
--cov -c The covariate matrix
```
## Inputs
## Gene counts
The script expects a gzipped file as an input for gene counts.
The input gene counts should be in the format:
```
Name, sample_id1, sample_id2..
```
Where Name is a column that has the gene ID (such as ENSG00000224533.11), which
should be in the same format as the gene IDs in the EQTL file
You can use the built in expression normalization method by passing --isnorm 0 when running the script.
You can also normalize expressions before feeding them into the script using DESEQ2 in R:
### Normalize counts using DESEQ2
```R
#import library
library(DESeq2)
#read in table
cts <- rix(read.csv2("example_table.tsv",sep='\t'))
#Set all extra columns other than the counts to NULL
cts$gene_id <- NULL
cts$Names <- NULL
#Delete original column names
colnames(cts) <- NULL
#Read dataset in
dds <- DESeqDataSetFromMatrix(countData = data.matrix(cts), colData = rep.int(1,ncol(cts)) , design = data.matrix(rep.int(1,ncol(cts))))
#Do the transform
dds <- DESeq(dds)
dds <- estimateSizeFactors(dds)
#Get the normalized counts
normalized_counts <- counts(dds, normalized=TRUE)
#Dump the result
write.csv(normalized_counts, "result.csv", row.names=FALSE)
```
## Haplotypes (phasing data)
This should be a VCF file in the format:
```
#CHROM POS ID REF ALT QUAL FILTER INFO sample1 sample2...
```
where #CHROM is the chr # POS is the position, ID is the variant ID in the format chr1_13550_G_A_b38, REF is the
reference allele, ALT is the alternative allele. A tabix index needs to be generated for the gzipped vcf file:
```
bgzip vcf_file.vcf && tabix -p vcf vcf_file.vcf.gz
```
More on tabix under
## Eqtl list
This file should contain gene IDs, variant IDs that match - it can also contain other stuff, but it needs to contain at least these two columns:
```
gene_id variant_id other_stuff1 other_stuff2...
```
## Covariate matrix (optional)
This matrix should contain a feature (such as gender) for each individual. The matrix should be in the format num_individuals x num_features
```
ID individual_1 individual_2 individual_3...
```