Skip to content
Snippets Groups Projects
Select Git revision
  • main default protected
1 result
Blame Permalink
README.md 19.81 KiB

SyntenyLink :dna:

Table of Contents :books:

  1. Overview
  2. Requirements
  3. Installation
  4. Usage
  5. Main programs
  6. Contact
  7. License

Overview :book:

===========

The SyntenyLink package has six major components: the SyntenyLink algorithm allows users to handle reconstruct subgenomes of polyploid species more conveniently and to separate the set of genes belong to each subgenome in the organism with the aid of reference proteomes of polyploid species and related ancestor. :seedling:

Pipeline

All programs are executed using command line options on Linux systems or Mac OS. Usage or help information are well built into the programs. :computer:

All code is copiable, distributable, modifiable, and usable without any restrictions.

Requirements :tools:

=============

To use SyntenyLink, ensure you have the following requirements:

  • Python 3.9
  • biopython==1.80
  • ipython==8.3.0
  • matplotlib==3.5.2
  • numpy==1.21.5
  • pandas==1.4.2
  • seaborn==0.11.2
  • pickle
  • csv
  • os
  • math
  • sys
  • re
  • warnings
  • wandb

Installation :gear:

=============

  1. Clone this repository to your local machine:
git clone git@git.cs.usask.ca:qnm481/syntenylink.git
cd SyntenyLink
  1. Install the required dependencies:
pip install -r requirements.txt
  1. Reproduce all the experiments:
python3 main_script.py -i abc_synteny.success.colinear -g -m -n -gt abc_groundtruth.xlsx -c abc_synteny.all.chains -bl abc_blastn.blast

Usage :rocket:

=============

Utilize this repository to replicate our experiments and explore the functionalities of SyntenyLink. The codebase is organized to help you easily navigate through different components and reproduce our results.

The following is the list of executable programs

Main programs

=============

Main programs (in the Scripts folder)

  1. SyntenyLink_bf.pl
  2. SyntenyLink_st.pl
  3. SyntenyLink_mbp.py
  4. SyntenyLink_wg.py
  5. SyntenyLink_sb.py
  6. SyntenyLink_mn.py
  7. main_script.py
  8. SyntenyLink_acc.py
  9. gap_threshold_selection.py
  10. minimum_block_length_selection.py

SyntenyLink_bf

This program, detects homologs between two species with blastp and involves a filtering step focusing on bit score and e-value to remove noise following two criteria: an e-value threshold of less than 1e-20 and a ratio between the bit score and the highest bit score greater than 0.6.

- Usage Reads in a data file: abc.blast. The abc.blast file holds blast hits after running blastp between baseline species and polyploid species of interest:

BraA01g000010.3C    AT1G43860.1 74.194  124 14  3   80  186 231 353 9.61e-56    188

Here is a typical parameter setting for generating the abc.blast file:

$ makeblastdb -in ref_pep.fa -dbtype prot -out ref_pep

$ blastall -i query_pep.fasta -p blastp -d ref_pep -m 8 -e 1e-5 -F F -v 5 -b 5 -o abc.blast -a 4

Note: for blastp you need to use protein fasta files of baseline model and polyploid species of interest. After getting the output blast result update gene start loci and end loci adding row start values from the bed file of each species before using SyntenyLink_bf.py.

It is advised that to make SyntenyLink_bf generate more reasonable results, the number of BLASTP hits for a gene should be restricted to around top 5. When you have abc.blast ready, put them in the same folder. Then you can simply use:

$ ./SyntenyLink_bf.pl dir/abc.blast

- Output The execution of SyntenyLink_bf outputs one blast file abc_blast_filtered.txt, containing filtered blast hits as follows:

BraA01g000010.3C    AT1G43860.1 74.194  124 14  3   80  186 231 353 9.61e-56    188
BraA01g000010.3C    AT3G04630.1 66.087  115 21  4   194 297 165 272 7.06e-33    125
BraA01g000010.3C    AT3G04630.3 66.087  115 21  4   194 297 164 271 7.88e-33    125

SyntenyLink_st

This program uses the output after performing synteny analysis using DAGchainer to build a chain of syntenic genes and compute the score of each chain. The modified version of filtered results from blastp (abc_blast_filtered_modified.txt) and DAGchainer (abc_synteny.aligncoords) are used to generate a syntelog table in which the gene chains are incorporated into their corresponding chromosomes with redundant chains (likely due to gene duplications or tandem or segmental duplications in the polyploid genome) placed in the "overlap" column in the syntelog table.

Reads in data file for DAGchainer: abc_blast_filtered_modified.txt. The abc_blast_filtered_modified.txt file holds the modified abc_blast_filtered.txt file matching the input file format of DAGchainer:

A1  BraA01g000010.3C    2944    3050    Chr1    AT1G43860.1 16622247    16622597    9.61E-56    188
A1  BraA01g000010.3C    3058    3161    Chr3    AT3G04630.1 1259234 1259503 7.06E-33    125

Note: After getting the abc_blast_filtered.txt file update query start, query end, subject start and subject end values incoorperating gene locus data from bed files of baseline species and polyploid species of interest. Then add chromosome which each gene belongs to in the bed file of each species before using DAGchainer. In order to do this, you can use transform_blast_to_dagchainer.py python script as follows:

$ python3 dir/transform_blast_to_dagchainer.py dir/abc_blast_filtered_modified.txt dir/query.bed (or dir/query.gff3) dir/subject.bed (or dir/subject.gff3)

When you run above python script it will generate an output file named transformed_blast_output_with_selected_columns.blast. Use this for dagchainer step.

Here is a typical parameter setting for generating the abc_synteny.aligncoords file:

$ ./run_DAG_chainer.pl -i dir/transformed_blast_output_with_selected_columns.blast -s -I

The abc_synteny.aligncoords file holds pairwise synteny blocks after running DAGchainer :

## alignment A1 vs. Chr1 Alignment #1  score = 5177.6 (num aligned pairs: 121):
A1  BraA01g026830.3C    17161339    17161504    Chr1    AT1G56580.1 21198405    21198568    2.180000e-109   50
A1  BraA01g026890.3C    17191267    17191318    Chr1    AT1G57550.1 21312544    21312593    3.270000e-24    40
A1  BraA01g026900.3C    17196325    17196618    Chr1    AT1G57610.3 21337612    21337818    8.840000e-106   84

- Usage Reads in two data files: abc_blast_filtered_modified.txt & ref_genelist.txt. The ref_genelist.txt file holds gff file like details of baseline species :

AT1G01010   AT1G01010.1 429 Chr1_1  Chr1    1   3631    5899    AT1G01010.1  NAC domain containing protein 1 
AT1G01020   AT1G01020.1 245 Chr1_2  Chr1    2   5928    8737    AT1G01020.1  Arv1-like protein 
AT1G01030   AT1G01030.1 358 Chr1_4  Chr1    4   11649   13714   AT1G01030.1  AP2/B3-like transcriptional factor family protein

To run SyntenyLink_st.pl you can simply use:

$ perl SyntenyLink_st.pl -d abc_synteny.aligncoords -g ref_genelist.txt

- Output The execution of SyntenyLink_st outputs four output files: abc_synteny.success.colinear, abc_synteny.failed.colinear, abc_synteny.chains.passed & abc_synteny.all.chains.

The abc_synteny.success.colinear file holds the main output file with the generated syntelog table in which the gene chains are incorporated into their corresponding chromosomes with redundant chains (likely due to gene duplications or tandem or segmental duplications in the polyploid genome) placed in the "overlap" column in the syntelog table:

BraA01g000010.3C    AT1G43860.1 74.194  124 14  3   80  186 231 353 9.61e-56    188
BraA01g000010.3C    AT3G04630.1 66.087  115 21  4   194 297 165 272 7.06e-33    125
BraA01g000010.3C    AT3G04630.3 66.087  115 21  4   194 297 164 271 7.88e-33    125

The abc_synteny.failed.colinear file holds the removed chains following the condition if the number of remaining colinear pairs are less than 6:

A6_Chr1_4   Chr1    AT1G14070   BraA06g010220.3C
A6_Chr1_4   Chr1    AT1G14080   BraA06g010230.3C
A6_Chr1_4   Chr1    AT1G14100   x
A6_Chr1_4   Chr1    AT1G14110   x
A6_Chr1_4   Chr1    AT1G14120   x
A6_Chr1_4   Chr1    AT1G14130   BraA06g010280.3C
A6_Chr1_4   Chr1    AT1G14140   x
A6_Chr1_4   Chr1    AT1G14150   x
A6_Chr1_4   Chr1    AT1G14160   x
A6_Chr1_4   Chr1    AT1G14170   x
A6_Chr1_4   Chr1    AT1G14180   x
A6_Chr1_4   Chr1    AT1G14185   x

The abc_synteny.chains.passed file holds the ids of set of passed chains with collinear pairs greater than 6:

A7_Chr1_1
A8.r_Chr1_1
A9.r_Chr1_1
A2_Chr1_1
A6_Chr1_1
A7.r_Chr1_1

The abc_synteny.all.chains file holds the all chains identified from DAGchainer:

A1_Chr1_1   A1  BraA01g026830.3C    17161339    17161504    Chr1    AT1G56580.1 21198405    21198568    2.180000e-109   50
A1_Chr1_1   A1  BraA01g026890.3C    17191267    17191318    Chr1    AT1G57550.1 21312544    21312593    3.270000e-24    40
A1_Chr1_1   A1  BraA01g026900.3C    17196325    17196618    Chr1    AT1G57610.3 21337612    21337818    8.840000e-106   84

SyntenyLink_mbp

This program identifies the main breakpoints of translocations in the abc_synteny.success.colinear file. It is is designed to detect fractionation gaps larger than a gap threshold. Synteny blocks capped by two breakpoints are grouped as a "super-synteny block" where gaps smaller than the gap threshold could exist within the super block. The input to this algorithm includes two parameters gap threshold, a minimum block length, and a data file syntelog table generated from Step 2, where collinear blocks are placed into the chromosomes of the organism. Next, the gene density of super-synteny blocks in each chromosome is calculated. This algorithm disregards densities below threshold density value. We selected a threshold density value of 0.1 in here. The chromosomes with gene density greater than 0.1 are ranked by the density of the blocks within the chromosome and assigned into candidate 𝑚 subgenomes, where 𝑚 denotes the ploidy level or the number of subgenomes to be reconstructed.

- Usage No input parameters or read in data files.

Note: You need to select the most suitable gap threshold and minimum block length for your data by running the gap_threshold_selection.py and minimum_block_length_selection.py scripts.

Here is a typical parameter setting for obtaining optimal gap threshold and block length values for you data:

$ python3 gap_threshold_selection.py -i abc_synteny.success.colinear

Then get the optimal gap threshold value from the console and run the following command:

$ python3 minimum_block_length_selection.py -i abc_synteny.success.colinear -g <output gap threshold value>

- Output Optimal gap threshold and block length values are printed to the console.

- Output The execution of SyntenyLink_mbp outputs two output files: Super_synteny_block_output.xlsx and Super_synteny_bl_sub_placement_density.xlsx.

The Super_synteny_block_output.xlsx file holds the generated super-synteny blocks before removing the noise taking density threshold into account:

Block no.   Block_start Block_end   Row start # Row end #   # genes in block    N1  N1.r    N2  N2.r    N3  N3.r    N4  N4.r    N5  N5.r    N6  N6.r    N7  N7.r    N8  N8.r
1   AT1G01010   AT1G01560   0   57  58  0.0 0.0 0.0 0.3684210526315789  0.0 0.6666666666666666  0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
2   AT1G01570   AT1G02205   58  123 66  0.0 0.0 0.43283582089552236 0.0 0.0 0.4626865671641791  0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
3   AT1G02210   AT1G14900   124 1491    1368    0.005113221329437546    0.0 0.3951789627465303  0.005843681519357195    0.3915266617969321  0.0 0.0577063550036523  0.1891891891891892  0.0 0.0 0.07596785975164354 0.26004382761139516 0.0 0.0 0.0 0.0

The Super_synteny_bl_sub_placement_density.xlsx file holds the subgenome placements of super-synteny blocks based on gene density, after removing the noise taking density threshold into account:

Block no.   Block_start Block_end   Row start # Row end #   # genes in block    N1  N1.r    N2  N2.r    N3  N3.r    N4  N4.r    N5  N5.r    N6  N6.r    N7  N7.r    N8  N8.r    N9  N9.r    Non_zero    subgenome1  subgenome2  subgenome3
1   AT1G01010   AT1G11860   0   1161    1162    0   0   0   0   0   0   0   0   0.544358312 0   0   0   0   0   0.354005168 0.308354866 0   0   3   N5  N8  N8.r
2   AT1G11870   AT1G13420   1162    1328    167 0   0   0   0   0   0   0   0   0   0.523809524 0   0   0   0   0.386904762 0.43452381  0   0   3   N5.r    N8.r    N8
3   AT1G13430   AT1G19470   1329    1957    629 0   0   0   0   0   0   0   0   0.598412698 0   0   0   0   0   0.33968254  0.422222222 0   0   3   N5  N8.r    N8

SyntenyLink_wg

This program uses a weighted direct graph to dynamically link blocks produced from Step 3 that are most likely to be in a subgenome using a combined information of fractionation and substitution patterns as well as continuity of gene chains.

- Usage Reads in two data files: abc_synteny.all.chains and abc_blastn.blast.

Note: You need to generate abc_blastn.blast file by running nucleotide blast using cds fasta files of baseline species and polyploid species of interest.

The abc_blastn.blast file holds blast hits after running blastn between baseline species and polyploid species of interest:

BraA01g000010.3C    AT1G43860.1 78.706  371 25  6   239 558 692 1059    7.94e-91    335

Here is a typical parameter setting for generating the abc_blastn.blast file:

$ makeblastdb -in ref_cds.fa -dbtype prot -out ref_cds

$ blastall -i query_cds.fasta -p blastn -d ref_cds -m 8 -e 1e-5 -F F -v 5 -b 5 -o abc_blastn.blast -a 4

- Output The execution of SyntenyLink_wg outputs number of output files matching the number of subgenomes in the species of interest: Super_synteny_graph_nodes_sub{k}.xlsx. There will be m number of files. k represents the corresponding subgenome number.

The Super_synteny_graph_nodes_sub{k}.xlsx file holds the nodes belong to each subgenome after traversing the graph:

Row start # Row end #   subgenome1  subgenome2  subgenome3
0   123 N10.r   N9  N8.r
124 698 N10 N8.r    N9.r
699 1029    N6  N9.r    N8.r

SyntenyLink_sb

This program retrieves genes "hidden" in small blocks that are missed in previous steps. Specically, we consider the chromosome blocks with densities lower than d<8= that are removed in previous steps in the generation of small blocks because these blocks may be more likely to exhibit ipping of synteny blocks between the forward and reverse strands resulting from segmental duplication. Following the generation of small blocks, small synteny blocks are incorporated into the corresponding super-synteny blocks, taking into account the subgenome placement of the super-synteny blocks as a reference.

- Usage No read in files or parameters.

- Output The execution of SyntenyLink_sb outputs number of output files: modified_chr_names{k+1}.xlsx, subgenome_placement_blocks.all{k+1}.xlsx, subgenome_placement_blocks.all.xlsx. k represents the corresponding subgenome number.

The modified_chr_names{k+1}.xlsx file holds the replaced gene id's with there corresponding chromosome names which they bellong to if they are not already placed inside the subgenome columns (to detect the genes that has been removed based on density threshold):

Gene_id	N1	N1.r	N2	N2.r	N3	N3.r	N4	N4.r	N5	N5.r	N6	N6.r	N7	N7.r
AT1G01900	0	0	0	0	0	0	0	0	0	1	0	0	0	0
AT1G01910	0	0	0	0	0	0	0	0	0	0	0	1	0	0
AT1G01920	0	0	0	0	0	0	0	0	0	0	0	1	0	0
AT1G01930	0	0	0	0	0	0	0	0	0	0	0	1	0	0
AT1G01940	0	0	0	0	0	0	0	0	0	0	0	1	0	0
AT1G01950	0	0	0	0	0	0	0	0	0	0	N6	0	0	0
AT1G01960	0	0	0	0	0	0	0	0	0	0	0	1	0	0
AT1G01970	0	0	0	0	0	0	0	0	0	1	0	0	0	0

The subgenome_placement_blocks.all{k+1}.xlsx file holds the placement of genes in subgenomes optimal for each subgenome considering the subgenome separation optimal for each subgenome in step 4:

Row start # Row end #   subgenome1  subgenome2  subgenome3
0   123 N10.r   N9  N8.r
124 698 N10 N8.r    N9.r
699 1029    N6  N9.r    N8.r

The subgenome_placement_blocks.all.xlsx file holds the final placement of genes in subgenomes in step 5:

Row start # Row end #   subgenome1  subgenome2  subgenome3
0   123 N10.r   N9  N8.r
124 698 N10 N8.r    N9.r
699 1029    N6  N9.r    N8.r

SyntenyLink_mn

Aimed to optimize the subgenome placements for each block (a row in the subgenome matrix) based on neighbourhood considerations in a subgenome, maximize the number of continuous blocks/rows from the same chromosome. The algorithm utilizes two window sizes, up and down, which determine the number of blocks above and below a given block, to define a neighbourhood.

- Usage No read in files Two input parameters: wup wdwn.

Note: You need to select the most suitable window up and window down parameters.

- Output The execution of SyntenyLink_sb outputs number of output files: Final_subgenome_placement_result.xlsx, Final_result.xlsx.

The Final_subgenome_placement_result.xlsx file holds the final placement of chromosome blocks in subgenomes:

Row start # Row end #   subgenome1  subgenome2  subgenome3
0   75  N10.r   N9  N8.r
76  78  N10.r   N9.r    N8.r
79  123 N10.r   N9  N8.r
124 698 N10 N9.r    N8.r
699 1035    N6  N9.r    N8.r
1036    1036    N6  N9  N8.r

The Final_result.xlsx file holds the final placement of genes inside blocks in subgenomes:

subgenome1  subgenome2  subgenome3
BraA10g000880.3C    x   x
BraA10g000870.3C    x   x
BraA10g000860.3C    x   x
BraA10g000850.3C    x   x
BraA10g000830.3C    BraA09g065940.3C    x
BraA10g000820.3C    x   x
x   BraA09g065960.3C    x
BraA10g000790.3C    x   x
BraA10g000780.3C    BraA09g065970.3C    x
BraA10g000770.3C    BraA09g065980.3C    x

main_script

This holds the main script that runs all the above scripts in order. It takes in the following parameters: -i input_file -g gap_threshold -m minimum_block_length -n number_of_subgenomes -gt ground_truth_file -c chains_file -bl blastn_file

Note: Before running main_script.py, you need to run the SyntenyLink_bf.pl and SyntenyLink_st.pl scripts

To run SyntenyLink_sb.py you can simply use:

$ python3 main_script.py -i abc_synteny.success.colinear -g -m -n -gt abc_groundtruth.xlsx -c abc_synteny.all.chains -bl abc_blastn.blast

- Output The execution of main_script outputs all the output files of the above scripts in the same directory as the input file.

SyntenyLink_acc

This holds the script that calculates the accuracy of the final placement of genes in subgenomes when there is a ground truth file.

- Usage One read in file: abc_groundtruth.xlsx

To run SyntenyLink_sb.py you can simply use:

$ python3 SyntenyLink_acc.py abc_groundtruth.xlsx

- Output The execution of SyntenyLink_sb prints the subgenome placemnet accuracy of each subgenome:

Exact match percentage for subgenome1: 84.15%
Exact match number for subgenome1: 11490
Missing genes for subgenome1: 2164
Exact match percentage for subgenome2: 61.61%
Exact match number for subgenome2: 6062
Missing genes for subgenome2: 3778
Exact match percentage for subgenome3: 65.30%
Exact match number for subgenome3: 5635
Missing genes for subgenome3: 2995

Contact :mailbox_with_mail:

===============

For any questions or inquiries, please feel free to open an issue on our repository or contact us at qnm481@usask.ca.

License :scroll:

===============

This project is licensed under the MIT License