Tutorials

BioKIT can be used for a multitude of different types of analyses. Documentation here provides a step-by-step outline for how to conduct some of these analyses.

  1. summarize genome assembly metrics,

  2. summarize multiple sequence alignment properties,

  3. calculate relative synonymous codon usage (RSCU), and

  4. calculate a novel metric for estimating codon optimization, gene-wise RSCU (or gw-RSCU).


1. Summarize genome assembly metrics

BioKIT implements numerous functions that can be used to evaluate diverse properties of a genome assembly.

In this tutorial, we will examine the contig-level genome assembly of Escherichia coli strain ATCC 11775, which was obtained from NCBI in September of 2021.

Download test data: E. coli strain ATCC 11775 genome assembly

To calculate all genome assembly metrics, use the following command:

biokit genome_assembly_metrics GCA_000734955.1_ASM73495v1_genomic.fna

4980585 Assembly size
14      L50
40      L90
124187  N50
33720   N90
0.5062  GC content
135     Number of scaffolds
93      Number of large scaffolds
4968723 Sum length of large scaffolds
382114  Longest scaffold
0.247   Frequency of A
0.2468  Frequency of T
0.2525  Frequency of C
0.2537  Frequency of G

If all metrics are not needed, individual metrics can be calculated. For example, calculate N50 of the genome assembly using the following command:

biokit n50 GCA_000734955.1_ASM73495v1_genomic.fna

124187

2. Summarize multiple sequence alignment properties

The properties of multiple sequence alignments are commonly reported in studies. To demonstrate how to calculate diverse properties of multiple sequence alignments, we will use the following single gene alignment from Steenwyk et al. 2019, mBio.

Download test data: Steenwyk_etal_mBio_2019_EOG091N44MS.aln.fa

To calculate diverse properties of the multiple sequence alignment, use the following command:

biokit alignment_summary Steenwyk_etal_mBio_2019_EOG091N44MS.aln.fa

General Characteristics
=======================
90      Number of taxa
624     Alignment length
517     Parsimony informative sites
555     Variable sites
69      Constant sites

Character Frequencies
=====================
T       8527
G       14835
C       15482
A       12966
-       4350

If all metrics are not needed, individual metrics can be calculated. For example, the number of constant sites can be calculated using the following command:

biokit constant_sites Steenwyk_etal_mBio_2019_EOG091N44MS.aln.fa

69

3. Calculate relative synonymous codon usage bias

Patterns of codon usage can be examined to reveal biases favoring some codons over others. In this section, we will calculate patterns of relative synonymous codon usage (RSCU) among genes encoded in the genome of Saccharomyces cerevisiae, the model budding yeast.

Download test data: GCA_000146045.2_R64_cds_from_genomic.fna

To calculate RSCU, use the following command:

biokit rscu GCA_000146045.2_R64_cds_from_genomic.fna

AGA     2.8425
GGU     1.8179
UUG     1.6747
UUA     1.6645
CCA     1.6281
UCU     1.5608
GUU     1.5466
GCU     1.4773
UAA     1.4221
GAA     1.4014
AUU     1.3807
CAA     1.371
ACU     1.3705
GAU     1.3024
CAU     1.2838
AGG     1.2789
UCA     1.2736
UGU     1.2424
CCU     1.2412
ACA     1.2315
AAU     1.1923
UUU     1.1892
GCA     1.1865
AAA     1.1684
UAU     1.1331
AUG     1.0
UGG     1.0
AGU     0.9759
UCC     0.9403
GGA     0.903
UGA     0.888
GCC     0.8851
GUA     0.8705
UAC     0.8669
CUA     0.8549
CGU     0.849
ACC     0.8452
AUA     0.8406
AAG     0.8316
UUC     0.8108
GUC     0.8093
AAC     0.8077
GGC     0.7898
CUU     0.7804
AUC     0.7787
GUG     0.7736
UGC     0.7576
CAC     0.7162
GAC     0.6976
UAG     0.6899
CUG     0.6745
AGC     0.6667
CCC     0.6332
CAG     0.629
GAG     0.5986
UCG     0.5827
ACG     0.5528
CCG     0.4974
GGG     0.4892
GCG     0.4512
CGA     0.4217
CGC     0.3587
CUC     0.351
CGG     0.2491

Values greater than 1 are reflective of codons that are over represented whereas values less than 1 are reflective of codons that are under represented and values equal to one occur as expected under a neutral expectation. Here, AGA is the most over represented codon and CGG is the most under represented codon.


4. Estimating codon optimization

Estimates of codon optimization can be used to identify genes that are expressed at high levels and are important for organismal ecology. Here, we will calculate gw-RSCU, a measure of codon optimization, in the coding sequences of Saccharomyces cerevisiae, the model budding yeast.

Download test data: GCA_000146045.2_R64_cds_from_genomic.fna

To calculate gw-RSCU, use the following command:

biokit gw_rscu GCA_000146045.2_R64_cds_from_genomic.fna

lcl|BK006935.2_cds_DAA06918.1_1 1.1972  1.0     0.4678
lcl|BK006935.2_cds_DAA06919.1_2 1.0037  0.9515  0.3458
lcl|BK006935.2_cds_DAA06920.1_3 1.1878  1.1892  0.3918
lcl|BK006935.2_cds_DAA06921.1_4 1.1082  0.9759  0.3646
lcl|BK006935.2_cds_DAA06922.1_5 1.0679  1.0     0.3447
lcl|BK006935.2_cds_DAA06923.1_6 1.1587  1.1892  0.3668
lcl|BK006935.2_cds_DAA06924.2_7 1.1429  1.1892  0.5033
lcl|BK006935.2_cds_DAA06925.1_8 1.2349  1.2315  0.4547
lcl|BK006935.2_cds_DAA06926.1_9 1.1726  1.2315  0.3764
...                             ...     ...     ...

Here, only the first nine lines of the output are shown. The first column is the gene identifier, the second, third, and fourth columns are the mean, median, and standard deviation of RSCU values observed in the gene. Our analyses demonstrate that the mean gw-RSCU performs best to examine codon optimization. These metrics can also be used to examine gene-wise codon usage biases.