Sina Majidian sinamajidian

Code is from Deep Learning AI course How Transformer LLMs Work

Set up

conda create -n llm python=3.12
conda activate llm
conda install conda-forge::transformers
conda install conda-forge::jupyterlab  
conda install pytorch::pytorch

In our Movi color study, we ran metabuli using the following script. We used two read datasets from CAMI (long and short read) for metagenomic classification.

Install

conda create -n metab python=3.12
conda activate metab

wget https://mmseqs.com/metabuli/metabuli-linux-avx2.tar.gz

import pyham
import logging

treeFile=fastoma_out+'/species_tree_checked.nwk'
orthoxmlFile=fastoma_out+'/FastOMA_HOGs.orthoxml'

logging.basicConfig(format='%(asctime)s %(levelname)-8s %(message)s', level=logging.INFO, datefmt='%Y-%m-%d %H:%M:%S')

PhylogeneticTree with FastOMA

Updating codes frm the F1000 paper on PhylogeneticTree on github for FastOMA

Impute-first pipeline for variant calling

This is bash script is based on the Impute-First github and the preprint.

Inputs:

Reference genome
HGSVC2 Reference panel
PLINK genetic mac
Novaseq HG002 sequencing reads

I faced with this error several times and searching on the net only results in using sudo apt-get.

/usr/bin/ld: cannot find -lcurl
/usr/bin/ld: cannot find -lbz2
collect2: error: ld returned 1 exit status

A diploid genome using SURVIVOR

Using the following bash code, you can create a diploid genome using SURVIVOR. Finally, you will have three files :

Two fasta file: sim1.fasta and sim2.fasta.
Truely phased vcf file: sim_e_merg.vcf.

Some lines of the intermediate files:

A simpe code for extracting the regions with high coverage (>depth_treshold) from a bed file. The input of the code is a bed file containing rows of regions with its depth as following format

21	9424000	9424500	69.00
21	9424500	9425000	69.00
21	9425000	9425500	67.00
21	9425500	9426000	66.00
21	9426000	9426500	65.00
21	9426500	9427000	66.00

	#!/usr/bin/python3

	import numpy as np
	from sys import argv


	file_fq_input_addrss = argv[1]
	file_fq_output_addrss = argv[2]

	file_fq_input= open(file_fq_input_addrss,'r');

	#!/usr/bin/env python3
	"""Coverage mean and standard deviation of autosomes

	Estimate the mean and standard deviation from a mosdepth coverage BED for
	positions with coverage in the range (0, 2 * non-zero mode). This estimate
	behaves well for PacBio HiFi WGS of human germline aligned to either hs37d5 and
	GRCh38, and may be useful for other situations as well.

	$ bash mosdepth --threads 3 --no-per-base --by 500 -m "${BAM%.*}.median" "${BAM}"
	$ tabix ${BAM%.*}.median.regions.bed.gz {1..22} \| python depth_mean_stddev.py