4.5.3 Bulk RNA seq

tresor.gene.simu_umi_len is a Python function in charge of simulating reads with respect to a series of UMI lengths at the bulk RNA-seq level.

Usage

We take the following command as an example to generate FastQ files with UMIs of lengths from 7 to 36.

PythonShell

import tresor as ts

for perm_i in range(1):
    print(perm_i)
    ts.gene.simu_umi_len(
        # initial sequence generation
        gspl=gspl,

        len_params={
            'umi': {
                'umi_unit_pattern': 3,
                'umi_unit_lens': np.arange(7, 36 + 1, 1),
            },
            'seq': 100,
        },
        material_params={
            'fasta_cdna_fpn': to('data/Homo_sapiens.GRCh38.cdna.all.fa.gz'),  # None False
        },
        seq_num=50,
        working_dir=to('data/simu/docs/'),

        condis=['umi', 'seq'],
        sim_thres=3,
        permutation=0,

        # PCR amplification
        ampl_rate=0.9,
        err_route='sptree',  # bftree sptree err1d err2d mutation_table_minimum mutation_table_complete
        pcr_error=1e-4,
        pcr_num=10,
        err_num_met='nbinomial',

        # PCR amplification
        seq_error=0.01,
        seq_sub_spl_number=200, # None
        # seq_sub_spl_rate=0.333,

        use_seed=True,
        seed=1,

        verbose=False,  # True False
        mode='short_read',  # long_read short_read

        sv_fastq_fp=to('data/simu/docs/'),
    )

tresor umilen_gene \
-cfpn ./tresor/data/amplrate_sl.yml \
-snum 50 \
-permut 0 \
-sthres 3 \
-wd ./tresor/data/simu/ \
-md short_read \
-is True \
-vb True

Attributes

Illustration

PythonShell

Attribute	Description
seq_num	number of RNA molecules. 50 by default
len_params	lengths of different components of a read. Lengths of UMIs vary in a range
seq_params	sequences of different components of a read, It allows users to add their customised sequences
material_params	a Python dictionary. Showing if cDNA libraries are provided, please use key word fasta_cdna_fpn. The human cDNA library can be downloaded through the Ensembl genome database
ampl_rate	float number ranging from 0 to 1
err_route	the computational algorithm to generate errors. There are 6 methods, including bftree, sptree, err1d, err2d, mutation_table_minimum, and mutation_table_complete.
pcr_error	PCR error rate
pcr_num	number of PCR cycles to amplify reads
err_num_met	the method to generate errors, that is, binomial or nbinomial
seq_error	sequencing error rate
seq_sub_spl_number	number of subsampling PCR amplified reads. It exists when seq_sub_spl_rate is specified to None
seq_sub_spl_rate	rate of subsampling PCR amplified reads. It exists when seq_sub_spl_number is specified to None
sv_fastq_fp	folder to save FastQ files
is_seed	if seeds are used to simulate sequencing libraries. This is designed to make in silico experiments reproducible
working_dir	working directory where all simulation results are about to be saved
condis	names of components that a read contains. It can contains an unlimited number of read components
sim_thres	similarity threshold. 3 by default
permutation	permutation times
mode	long_read or short_read
verbose	whether to print intermediate results

Attribute	Description
cfpn	location to the yaml configuration file. Users can specify the atrributes illustrated on the Python tab in the .yml file.
snum	number of sequencing molecules
permut	permutation times
sthres	similarity threshold. 3 by default
wd	working directory where all simulation results are about to be saved
md	long_read or short_read mode
is	if seeds are used to simulate sequencing libraries. This is designed for reproducible in silico experiments
vb	whether to print intermediate results

Output

Console

======>simulation completes in 2.8730275630950928s
======>simulation completes in 2.9519758224487305s
======>simulation completes in 2.930018663406372s
======>simulation completes in 3.0289931297302246s
======>simulation completes in 3.19893479347229s
======>simulation completes in 3.496039390563965s
======>simulation completes in 3.4660286903381348s
======>simulation completes in 3.4059395790100098s
======>simulation completes in 3.371000051498413s
======>simulation completes in 3.4959704875946045s
======>simulation completes in 3.7300221920013428s
======>simulation completes in 3.6379969120025635s
======>simulation completes in 3.91005277633667s
======>simulation completes in 4.011698961257935s
======>simulation completes in 4.11646032333374s
======>simulation completes in 4.311992883682251s
======>simulation completes in 4.139020919799805s
======>simulation completes in 4.244063138961792s
======>simulation completes in 4.463680982589722s
======>simulation completes in 4.64800238609314s
======>simulation completes in 4.433006048202515s
======>simulation completes in 5.417316675186157s
======>simulation completes in 4.551185369491577s
======>simulation completes in 4.674246311187744s
======>simulation completes in 5.081998348236084s
======>simulation completes in 4.961503744125366s
======>simulation completes in 5.0400168895721436s
======>simulation completes in 5.126001596450806s
======>simulation completes in 4.928617477416992s
======>simulation completes in 5.1075379848480225s
Finished!

Understanding files

The resultant files of the simulated reads are shown as follows.

Fig 1. Generated FastQ files

It is worth mentioning the libraries of UMIs with different lengths rendered in the following form.

Fig 2. Sequencing libraries