- 2023-01-21 20:00, Zhengyang Wen
- This is a repository for data and scripts of LOST & FOUND
LOST & FOUND (LOcal Sequence based Tracing Founctional Orthologous UNit Death) is a pipeline for gene loss identification. With orthologous relations and genome alignment data, it can identify gene loss events efficiently. For more detailed information, please check the manuscript Genome-wide identification of gene loss events suggests loss relics as a potential source for functional lncRNAs in human.
Dependencies:
Scripts: python package - pyinterval
linux command - bedtools
Data: perl package - Ensembl Perl API
tom@linux$ git clone git@github.com:gao-lab/LOST-and-FOUND.git
Here, we demostrate how to use LOST & FOUND. The example used here is the identification of human gene loss in the manuscript Genome-wide identification of gene loss events suggests loss relics as a potential source for functional lncRNAs in human.
Please note that to apply LOST & FOUND in other speices, the user may need to provide orthologous relation and genome alignment data in required format. We will show how to acquire required data in the Data Process part.
This step is to generate gene loss candidates of orthologous groups.
It requires the following input:
--IDheader
# Headers of Ensembl Gene ID and their corresponding species name written in “Python dict format”
--OrthologousPath
# Path that contains orthologous relation files. It should contains the orthologous relation files between anchor and target species branch.
The orthologous relations file should contain the following columns (delimited by tab):
#Gene stable ID Chromosome/scaffold name Gene start (bp) Gene end (bp) Human gene stable ID Human chromosome/scaffold name Human chromosome/scaffold start (bp) Human chromosome/scaffold end (bp) Human homology type %id. target Human gene identical to query gene %id. query gene identical to target Human gene.
Orthologous relation files can be obtained via Ensembl BioMart.
--AnchorSpeciesSet
# Names of anchor species written in “Python list format”
--TargetTree
# The phylogeny tree of target branch used for maximum parsimony inference, written in “Python list format”
--TargetSpecies
# Target species name
--OutputPath
# Path for output results
The output of this step is Loss_candidates_of_orthologous_group. Each line of this output represents an orthologous group and it consists of loss node, gene state in each species, gene number in each species and its gene members.
Here, we show an example in example/result/Loss_candidates_of_orthologous_group:
Loss_node:Macaca-Gorilla-Human-Chimp Exist_Shrew_1_Gorilla_0_Rat_1_Ryukyu_1_Human_0_Macaca_1_Chimp_1_Mouse_1 Num_Shrew_1_Gorilla_0_Rat_1_Ryukyu_1_Human_0_Macaca_1_Chimp_1_Mouse_1 MGP_PahariEiJ_G0025857 ENSMMUG00000016490 ENSPTRG00000045117 ENSMUSG00000027443 ENSRNOG00000004946 MGP_CAROLIEiJ_G0024414
Loss_node:Human-Chimp Exist_Shrew_1_Gorilla_1_Rat_1_Ryukyu_1_Human_0_Macaca_1_Chimp_1_Mouse_1 Num_Shrew_1_Gorilla_1_Rat_1_Ryukyu_1_Human_0_Macaca_1_Chimp_1_Mouse_1 ENSPTRG00000051474 ENSMUSG00000046828 ENSMMUG00000000353 MGP_PahariEiJ_G0027346 ENSRNOG00000022762 MGP_CAROLIEiJ_G0014103 ENSGGOG00000026202
Loss_node:Macaca-Gorilla-Human-Chimp Exist_Shrew_1_Gorilla_0_Rat_1_Ryukyu_1_Human_0_Macaca_1_Chimp_0_Mouse_1 Num_Shrew_3_Gorilla_0_Rat_7_Ryukyu_4_Human_0_Macaca_1_Chimp_0_Mouse_4 MGP_PahariEiJ_G0025265 ENSMUSG00000075159 MGP_PahariEiJ_G0025266 ENSMUSG00000075161 ENSMUSG00000075160 ENSRNOG00000050825 MGP_PahariEiJ_G0025264 ENSRNOG00000048383 MGP_CAROLIEiJ_G0023816 ENSRNOG00000050876 MGP_CAROLIEiJ_G0023814 MGP_CAROLIEiJ_G0023815 ENSRNOG00000045627 ENSMMUG00000016148 ENSMUSG00000075163
ENSRNOG00000049276 ENSRNOG00000047728 MGP_CAROLIEiJ_G0023813 ENSRNOG00000048820
Loss_node:Human-Chimp Exist_Shrew_1_Gorilla_1_Rat_1_Ryukyu_1_Human_0_Macaca_1_Chimp_1_Mouse_1 Num_Shrew_1_Gorilla_1_Rat_1_Ryukyu_1_Human_0_Macaca_1_Chimp_1_Mouse_1 ENSMUSG00000036924 ENSRNOG00000005103 ENSMMUG00000031138 ENSGGOG00000038343 MGP_CAROLIEiJ_G0024416 ENSPTRG00000050976 MGP_PahariEiJ_G0025859
Loss_node:Macaca-Gorilla-Human-Chimp Exist_Shrew_1_Gorilla_0_Rat_1_Ryukyu_1_Human_0_Macaca_1_Chimp_0_Mouse_1 Num_Shrew_1_Gorilla_0_Rat_1_Ryukyu_1_Human_0_Macaca_1_Chimp_0_Mouse_1 ENSRNOG00000042494 ENSMMUG00000046571 MGP_CAROLIEiJ_G0028840 MGP_PahariEiJ_G0022598 ENSMUSG00000079346
Loss_node:Macaca-Gorilla-Human-Chimp Exist_Shrew_1_Gorilla_0_Rat_1_Ryukyu_1_Human_0_Macaca_1_Chimp_0_Mouse_1 Num_Shrew_1_Gorilla_0_Rat_1_Ryukyu_1_Human_0_Macaca_1_Chimp_0_Mouse_1 ENSRNOG00000036897 ENSMUSG00000050645 MGP_PahariEiJ_G0025907 MGP_CAROLIEiJ_G0024464 ENSMMUG00000008796
tom@linux$ cd bin
tom@linux$ bash Identify_loss_candidates.sh \
--IDheader '{"Mouse":"ENSMUSG","Rat":"ENSRNOG","Ryukyu":"MGP_CAROLIEiJ_G","Shrew":"MGP_PahariEiJ_G","Human":"ENSG","Chimp":"ENSPTRG","Macaca":"ENSMMUG","Gorilla":"ENSGGOG"}' \
--OrthologousPath ../example/data/OrtholougousRelation/ \
--AnchorSpeciesSet '["Mouse","Rat","Ryukyu","Shrew"]' \
--TargetTree '[[["Chimp","Human"],"Gorilla"],"Macaca"]' \
--TargetSpecies Human \
--OutputPath ../example/result/
This step is to confirm Partial Loss Events and tracing their loss relics.
It requires the following input:
--LossCandidate
# Loss candidate file, the output of last step
--AnchorHeader
# Header of Ensembl Gene ID of the species used as anchor tracing loss relics
--TargetSpecies
# Target species name. Please make sure it is identical to the target species name in the genome alignment data
--GenomeAlignment
# Path for genome alignment data. You can take `example/data/GenomeAlignment/Blocks_for_Loss_candidates_HUMAN` as an example.
In the Data Process part, we will show how to acquire this data
--RegionThreshold
# Threshold that determines whether merge or split regions when tracing loss relics
--OutputPath
# Path for output results
The output of this step includes Partial_Loss_Events and Orthologous_group_candidate_for_Complete_Loss. Partial_Loss_Events contains Partial Loss Events, each line of which represents a Partial Loss Event. And it is in table format (delimited by tab): #RelicChr RelicStart RelicEnd AnchorGene.
Here, we show an exmaple in example/result/Partial_Loss_Events:
20 50081127 50113258 ENSMUSG00000078923
3 130244494 130264171 ENSMUSG00000032572
18 35290366 35306215 ENSMUSG00000063281
12 6867383 6870974 ENSMUSG00000023456
2 55284026 55287618 ENSMUSG00000032673
1 225119276 225402030 ENSMUSG00000047369
1 11654670 11662291 ENSMUSG00000029001
11 57741398 57743051 ENSMUSG00000076437
2 200871146 200889344 ENSMUSG00000026035
3 101649455 101651246 ENSMUSG00000071533
Orthologous_group_candidate_for_Complete_Loss contains the candidates for Complete Loss Events.
tom@linux$ bash Identify_Partial_Loss.sh \
--LossCandidate ../example/result/Loss_candidates_of_orthologous_group \
--AnchorHeader ENSMUSG \
--TargetSpecies homo_sapiens \
--GenomeAlignment ../example/data/GenomeAlignment/Blocks_for_Loss_candidates_HUMAN \
--OutputPath ../example/result/ \
--RegionThreshold 500000
This step is to confirm Complete Loss Events and locate their synteny regions.
It requires the following input:
--AnchorHeader
# Header of Ensembl Gene ID of the species used as anchor
--TargetSpecies
# Target species name. Please make sure it is identical to the target species name in the genome alignment data.
--GenomeAlignment
# Path for nearby genome blocks data. You can take `example/data/GenomeAlignment/Blocks_for_Complete_Loss_HUMAN` as an example.
In the Data Process part, we will show how to acquire this data
--NearestBlocksAlignment
# Path for closest genome blocks data. You can take `example/data/GenomeAlignment/Blocks_for_tracing_Complete_Loss_region_HUMAN` as an example.
In the Data Process part, we will show how to acquire this data
--OutputPath
# Path for output results
The output of this step includes Complete_Loss_Events and Complete_Loss_synteny_region. Complete_Loss_Events contains the anchor gene id of each Complete Loss Event. Complete_Loss_synteny_region contains their synteny region and it is in table format (delimited by tab): #AnchorGene SyntenyChr SyntenyStart SyntenyEnd
Here, we show an exmaple in example/result/Complete_Loss_synteny_region:
ENSMUSG00000072647 12 111842236 112160327
ENSMUSG00000110012 11 4690793 5581774
ENSMUSG00000111273 12 55185714 55790565
ENSMUSG00000023577 3 51721810 51935219
ENSMUSG00000061295 11 48110031 49952593
ENSMUSG00000058981 1 159218919 159478969
ENSMUSG00000062527 1 159218919 159478969
ENSMUSG00000063251 17 41008954 41275722
ENSMUSG00000032493 3 46744650 46847719
ENSMUSG00000044041 17 41491422 41552084
ENSMUSG00000022798 3 119093920 119176274
ENSMUSG00000045031 4 122710737 122827273
ENSMUSG00000021953 8 11226515 11558020
ENSMUSG00000069708 6 132503457 132734611
ENSMUSG00000069707 6 132503457 132734611
tom@linux$ bash Identify_Complete_Loss.sh \
--AnchorHeader ENSMUSG \
--TargetSpecies homo_sapiens \
--GenomeAlignment ../example/data/GenomeAlignment/Blocks_for_Complete_Loss_HUMAN \
--NearestBlocksAlignment ../example/data/GenomeAlignment/Blocks_for_tracing_Complete_Loss_region_HUMAN \
--OutputPath ../example/result/
This step is to filter loss events overlapped with assembly gap.
It requires the following input:
--PartialLoss
# Partial Loss Events, the output of step 2
--CompleteLoss
# Complete Loss Events with synteny regions, the output of step 3
--AssemblyGap
# Assembly Gap file, in bed format
--OutputPath
# Path for output results
The output of this step includes Partial_Loss_Events_without_gap and Complete_Loss_Events_without_gap.
tom@linux$ bash Filter_assembly_gap.sh \
--PartialLoss ../example/result/Partial_Loss_Events \
--CompleteLoss ../example/result/Complete_Loss_synteny_region \
--AssemblyGap ../example/data/Human_assembly_gap.bed \
--OutputPath ../example/result/
In this part, we will show how to acquire the genome alignment data required in the pipeline. Please note that this part relies on the Ensembl Perl API. Make sure that it has been installed and can be used properly.
Note: Since the genome alignment dataset is large, it would be SLOW to acquire these data through Ensembl server. We strongly recommend you to install the Ensembl database and acquire these data locally.
This step is to acquire the geneome alignment data Blocks_for_Loss_candidates required in step 2.Identify Partial Loss Events.
It requires the following input:
--Host
# Host of the ensembl database. 'ensembldb.ensembl.org' to access database through Ensembl server
--User
# User of the ensembl database. 'anonymous' to access database through Ensembl server
--AnchorSpecies
# Anchor species name, make sure this name can be recognized by Ensembl Perl API
--MethodLinkType
# Method name for mulitple genome alignment dataset, such as "EPO", "PECAN" or "EPO_EXTENDED". You can visit Ensembl to check the available alignment dataset
--SpeciesSetName
# Species set name for mulitple genome alignment dataset, such as "mammals", "primates" or "amniotes". You can visit Ensembl to check the available alignment dataset
--AnchorIDHeader
# Header of Ensembl Gene ID of the species used as anchor
--LossCandidate
# Loss candidates dataset, the output of the first step 'Identify gene loss candidates'
--OrthologousPath
# Path that contains orthologous relation files
--OutputPath
# Path for output results
tom@linux$ cd bin
tom@linux$ bash Get_alignmentblocks_for_loss_candidates.sh \
--Host 'ensembldb.ensembl.org' \
--User 'anonymous' \
--AnchorSpecies Mouse \
--MethodLinkType EPO \
--SpeciesSetName mammals \
--AnchorIDHeader ENSMUSG \
--LossCandidate ../example/result/Loss_candidates_of_orthologous_group \
--OrthologousPath ../example/data/OrtholougousRelation/ \
--OutputPath YourDatapath/
This step is to acquire the geneome alignment data Blocks_for_tracing_Complete_Loss_region and Blocks_for_Complete_Loss required in step 3.Identify Complete Loss Events.
It requires the following input:
--Host
# Host of the ensembl database. 'ensembldb.ensembl.org' to access database through Ensembl server
--User
# User of the ensembl database. 'anonymous' to access database through Ensembl server
--AnchorSpecies Mouse
# Anchor species name, make sure this name can be recognized by Ensembl Perl API
--TargetSpecies homo_sapiens
# Target species name. Please make sure it is identical to the target species name in the genome alignment data
--MethodLinkType
# Method name for mulitple genome alignment dataset, such as "EPO", "PECAN" or "EPO_EXTENDED". You can visit Ensembl to check the available alignment dataset
--SpeciesSetName
# Species set name for mulitple genome alignment dataset, such as "mammals", "primates" or "amniotes". You can visit Ensembl to check the available alignment dataset
--AnchorIDHeader
# Header of Ensembl Gene ID of the species used as anchor
--CompleteLossCandidate
# Complete Loss candidates dataset, the output of the second step 'Identify Partial Loss Events'
--OrthologousPath
# Path that contains orthologous relation files
--OutputPath
# Path for output results
--ExtendLength
# Length extended in the upstream and downstream of anchor gene for alignment blocks. Here, we recommend the value
tom@linux$ bash Get_alignmentblocks_for_CompleteLoss.sh \
--Host 'ensembldb.ensembl.org' \
--User 'anonymous' \
--AnchorSpecies Mouse \
--TargetSpecies homo_sapiens \
--MethodLinkType EPO \
--SpeciesSetName mammals \
--AnchorIDHeader ENSMUSG \
--CompleteLossCandidate ../example/result/Orthologous_group_candidate_for_Complete_Loss \
--OrthologousPath ../example/data/OrtholougousRelation/ \
--OutputPath YourDatapath/ \
--ExtendLength 1000000