KinOrtho: a combination of full-length and domain-based orthology inference methods

Requirement

Please ensure the following software is installed:

• Python (v3.7.4 or later) https://www.python.org/downloads/
  • Instruction: https://devguide.python.org/
• NCBI BLAST+ (v2.9.0 or later) https://ftp.ncbi.nlm.nih.gov/blast/executables/LATEST/
  • Instruction: https://www.ncbi.nlm.nih.gov/books/NBK279690/
• MCL (v14.137 or later) https://micans.org/mcl/src/
  • Instruction: https://micans.org/mcl/index.html?sec_license

Installation

PIP is a package manager for Python packages. If you do not have PIP installed, please use the commands below:

curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py
python get-pip.py

To install KinOrtho, please use the command below:

pip install kinortho

NOTE: after installation, you can execute kinortho using "kinortho" instead of "python kinortho.py" shown in the Usage section below. For example, to identify the orthologs of human protein kinases based on both full-length and domain-based methods, you can use the command below:

kinortho -i ./example/reference_proteomes/ -f ./example/HumanProteinKinase.fasta -d ./example/HumanKinaseDomain.fasta

Usage

• Identify all orthologous relationships without query sequences

  • python kinortho.py -i <reference_proteomes> [OPTIONS]
    • Example:

     python kinortho.py -i ./example/reference_proteomes/
    

    • Output: https://github.com/esbgkannan/KinOrtho/tree/main/example/output/no-query/

• KinOrtho (Full-length): full-length orthology inference with query sequences

  • python kinortho.py -i <reference_proteomes> -f <full_length_query_seqs>
    • Example:

     python kinortho.py -i ./example/reference_proteomes/ -f ./example/HumanProteinKinase.fasta
    

    • Output: https://github.com/esbgkannan/KinOrtho/tree/main/example/output/full-length/

• KinOrtho (Domain-based): domain-based orthology inference with query sequences

  • python kinortho.py -i <reference_proteomes> -d <domain_based_query_seqs>
    • Example:

     python kinortho.py -i ./example/reference_proteomes/ -d ./example/HumanKinaseDomain.fasta
    

    • Output: https://github.com/esbgkannan/KinOrtho/tree/main/example/output/domain-based/

• KinOrtho (Overlapping): combines the full-length and domain-based orthology inference results

  • python kinortho.py -i <reference_proteomes> -f <full_length_query_seqs> -d <domain_based_query_seqs>
    • Example:

     python kinortho.py -i ./example/reference_proteomes/ -f ./example/HumanProteinKinase.fasta -d ./example/HumanKinaseDomain.fasta
    

    • Output: https://github.com/esbgkannan/KinOrtho/tree/main/example/output/overlapping/

• Convert KinOrtho's output into a file with QfO submission format

  • python toqfo.py <kinortho_output>
    • Example:

     python toqfo.py ./example/output/overlapping/Results_Overlapping.txt
    

    • Output: https://github.com/esbgkannan/KinOrtho/blob/main/example/output/overlapping/Results_Overlapping.rels.raw

Options

• -i <reference_proteomes>
  • The folder of input proteomes (required)
• -f <full_length_query_seqs>
  • Full-lengt query sequences (FASTA format)
• -d <domain_based_query_seqs>
  • Domain-based query sequences (FASTA format)
• -o <out_folder>
  • Output folder (default: ./kinortho_out/)
• -E <e_value>
  • E-value threshld (default: 1e-5)
• -t <num_threads>
  • Number of threads (default: 1)
• -I <inflation_value>
  • This value handles for affecting cluster granularity. (default: 1.5)
• -e <min_ev>
  • Minimal E-value. This value will replace the E-value '0' in BLAST output. (default: 1e-200)
• -s <start_step>
  • Start step. Value: {0, 1, 2, 3, 4, 5, 6}. (default: 0)
• -S <stop_step>
  • Stop step. Value: {0, 1, 2, 3, 4, 5, 6}. (default: 6)

Steps: 0: Initiation. 1: Homology Search. 2: Building BLAST DB. 3: All-vs-all Homology Search. 4: Orthology Inference. 5: Cluster Analysis. 6: Combining Results.

Output format

• KinOrtho (Full-length/No-query):

1. Species_1 - the proteome file name of protein 1
2. Protein_1 - the sequence ID of protein 1
3. Species_2 - the proteome file name of protein 2
4. Protein_2 - the sequence ID of protein 2
5. E-value - the E-value of BLAST result (between protein 1 and protein 2)
6. Weight - the weight between protein 1 and protein 2 in the graph
7. Relationship - orthologous relationship between protein 1 and protein 2 (Ortholog/In-paralog/Co-ortholog)

• KinOrtho (Domain-based):

1. Species_1 - the proteome file name of protein 1
2. Protein_1 - the sequence ID of protein 1
3. Domain_1 - the domain region of protein 1 (dom[INDEX][START][END])
4. Species_2 - the proteome file name of protein 2
5. Protein_2 - the sequence ID of protein 2
6. Domain_2 - the domain region of protein 2 (dom[INDEX][START][END])
7. E-value - the E-value of BLAST result (between domain 1 and domain 2)
8. Weight - the weight between domain 1 and domain 2 in the graph
9. Relationship - orthologous relationship between domain 1 and domain 2 (Ortholog/In-paralog/Co-ortholog)

• KinOrtho (Overlapping):

1. Species_1 - the proteome file name of protein 1
2. Protein_1 - the sequence ID of protein 1
3. Domain_1 - the domain region of protein 1 (dom[INDEX]-[START]-[END])
4. Species_2 - the proteome file name of protein 2
5. Protein_2 - the sequence ID of protein 2
6. Domain_2 - the domain region of protein 2 (dom[INDEX]-[START]-[END])
7. E-value_Full - the E-value of BLAST result (between protein 1 and protein 2)
8. E-value_Domain - the E-value of BLAST result (between domain 1 and domain 2)
9. Weight_Full - the weight between protein 1 and protein 2 in the full-length graph
10. Weight_Domain - the weight between domain 1 and domain 2 in the domain-based graph
11. Relationship - orthologous relationship between protein 1/domain 1 and protein 2/domain 2 (Ortholog/In-paralog/Co-ortholog)

Example datasets

• Reference proteomes (human, mouse, and fruit fly)
• Query sequences (full-length) (545 human protein kinases)
• Query sequences (domain-based) (558 human kinase domains)

Updates

• v1.0.0
  • The first version of KinOrtho

Citation

To cite our work, please refer to:

KinOrtho: a method for mapping human kinase orthologs across the tree of life and illuminating understudied kinases. Liang-Chin Huang, Rahil Taujale, Nathan Gravel, Aarya Venkat, Wayland Yeung, Dominic P Byrne, Patrick A Eyers, and Natarajan Kannan. Preprint