Genomic and transcriptomic evolution of D. suzukii

Table of contents

1. Context
2. Introduction to the project
3. Data
4. SNP Calling Workflow
5. Diversity Analysis
6. TE Analysis
7. Authors
8. References

1. Context

Since its recent invasion in the European and American continents, Drosophila suzukii, has become a major pest of berry crops. Contrary to other Drosophila species that develop themselves on damaged or rotten fruit, D. suzukii is able to lay its eggs in healthy fruit before harvest, using its sclerotinized ovipositor. In order to control the D. suzukii population it is therefore required to improve our knowledge about its ability to adapt to a local environment outside its origin areas.

A previous study focused on phenotipic changes during a selective experiment showed a temporal adaptation of D. suzukii populations in the three different fruit environments. This process produces a fitness improvement which corresponds to a greater ability to reproduce and could explain the adaptation abilities of D. suzukii. Based on the same experiment, a second study has been started by the LBBE to see if the D. suzukii genome and transcriptome are affected by the environment.

The project is focused on both genes and Transposable Elements (TE) expression in order to find the variants explaining the local environment adaptation of D. suzukii. A significant number of differentially expressed genes across generations has been found with RNAseq analysis between the G0 and the G7 generations. These variations seems to be independant of the fruit media. Moreover, in differentially expressed (DE) genes an increase of the number of Single Nucleotide Polymorphism (SNP) was found in G7 compared to G0, but this result was not expected. Natural selection generally keeps alleles which allow the best adaptation, and it is synonymous with a decrease in SNP number.

2. Introduction to the project

2.1 Polymorphism analysis

The goal here is to analyse D. suzukii polymorphism performing a variant calling on the whole genome (using poolseq data) in order to count the number of different genomic positions between G0 and G12, and between fruits media.
The position of the genetic variations will be compared with previous SNP analysis conducted on RNA-seq data to see if we also find an increase of the SNP number. To have a view more global on the genome's diversity, several diversity measures will be calculated: pi's , Watterson's , Tajima's D.

2.2 TE analysis

Transposable elements (TE) represent 47% of the D. suzukii genome and can be important in the adaptation process, interfering in gene regulation during evolution. The aim is to determine TEs age on our data and compute their abundance. Comparing abundance graphs between G0 and G12 will inform us if some recent TE insertions happened.

All the project objectives are presented more precisely in the book of specifications file.

3. Data

• DNA poolseq fastq from D.suzukii obtained with paired-end Illumina sequencing (2x100pb), with 40 individuals for each pool:
  • One poolseq for G0;
  • One poolseq for each media in G12 (strawberry, cherry and cranberry).

File name	Number of reads
G0-MTP	242 990 252
G12-cerise	214 319 262
G12-cranb	261 528 888
G12-fraise	216 028 586

Location: pedago-ngs

/localdata/pandata/students/M2_projet_15/data

• Partial assembly of D. suzukii genome as reference (4 chromosomes, 542 contigs);

Location: pedago-ngs

/localdata/pandata/students/M2_projet_15/ref/Drosophila-suzukii-contig.fasta

• Annotation file at the format gff3. Location: pedago-ngs

/localdata/pandata/students/M2_projet_15/ref/Drosophila-suzukii-annotation-3-ws.gff3

4. SNP Calling Workflow

4.1 Quality control

First of all we checked the quality of the data, using fastqc (version v0.11.9):

fastqc sample.fastq.gz

Location: pedago-ngs

/localdata/pandata/students/M2_projet_15/quality_control

Then, we merged results from fastqc using multiqc in the folder with fastqc reports. The html report is available in the quality_control folder.
We noticed a high GC rate in the reads, but it seems to be in accordance with the results obtained by other researchs on D. suzukii.

4.2 Mapping

1. Genome indexing:

bwa index \
    -p Drosophila-suzukii-contig.fasta.fai \
    -a bwtsw Drosophila-suzukii-contig.fasta

• -p: index name;
• -a: index algorithm (bwtsw for long genome).

2. Mapping samples on the reference:

bwa aln -t 14 Drosophila-suzukii-contig.fasta.fai sample_1.fastq.gz > sample_1.sai
bwa aln -t 14 Drosophila-suzukii-contig.fasta.fai sample_2.fastq.gz > sample_2.sai

• -t: number of threads.

3. Merge the .sai files of forward and reverse reads:

bwa sampe Drosophila-suzukii-contig.fasta.fai \
    sample_1.sai sample_2.sai \
    sample_1.fastq.gz sample_2.fastq.gz > sample_pe.sam

• sampe: take in account to have paired ends reads.

Mapping stats

Computation of mapped files statistics using:

samtools stats sample.sam

sample	raw sequences	mapped and paired squences
G0-MTP_pe	242495252	194326801
G12-fraise_pe	216028586	172898366
G12-cerise_pe	214319262	172440052
G12-cranb_pe	261528888	212135984

Conversion to .bam

samtools view -bS -@ 3 sample.sam > sample.bam

• -bS: output in the BAM format ignoring for compatibility with previous samtools version;
• -@: number of threads.

Location: pedago-ngs

/localdata/pandata/students/M2_projet_15/mapping_bwa

NB: Before using bwa, we tried to use hisat2 but, for an unkown reason it could not map the strawberry sample.

4.3 Variant Calling

The SNP calling on pooled data is still not largely used because it is very innovative. For this reason it has required several tests (with different tools) and a long time to be completed.
Finally, we used GATK to perform our SNP calling and for that we were inspired by M. Tabourin's workflow (Analyse experience d'évolution expérimentale D.suzukii) and from this document.

1. Sorting .bam files:

# GATK version: gatk-4.2.2.0
gatk SortSam -I sample.bam -O sample_sorted.bam -SO queryname -VALIDATION_STRINGENCY SILENT

• -I: input file;
• -O: output file;
• -SO: sort order of output file;
• -VALIDATION_STRINGENCY: Validation stringency for all SAM files read by this program. Setting stringency to SILENT can improve performance when processing a BAM file in which variable-length data (read, qualities, tags) do not otherwise need to be decoded.

2. Add read group info:

gatk AddOrReplaceReadGroups \
    -I sample_sorted.bam \
    -O sample_ReadGroups.bam \
    -RGID name \
    -RGLB name \
    -RGPL ILLUMINA \
    -RGPU name \
    -RGSM sample \
    -VALIDATION_STRINGENCY SILENT

• -I: input file;
• -O: output file;
• -RGID: Read-Group ID;
• -RGLB: Read-Group library;
• -RGPL: Read-Group platform;
• -RGPU: Read-Group platform unit(barcode);
• -RGSM: Read-Group sample name;
• -VALIDATION_STRINGENCY: Validation stringency for all SAM files read by this program.

NB: Read group information flags (-RG*) are mandatory, even if you don't have them.

3. Remove duplicates:

gatk MarkDuplicatesSpark \
    -I sample_ReadGroups.bam \
    -O sample_MarkDuplicated.bam \
    --spark-master local[16]

• -I: input file;
• -O: output file;
• --spark-master: to parallelize the work.

4. Detect variant sites in the dataset:

gatk HaplotypeCaller \
    -R reference.fasta \
    -I sample_MarkDuplicated.bam \
    --sample-ploidy 80 -O sample.vcf \
    --max-genotype-count 91881

• -R: genome reference;
• -I: input file;
• -O: output file (vcf format);
• --sample-ploidy: it allows to take in account of using pooled data:
  In our case we have a ploidy of 2 (D. suzukii is a diploïd organism) and a pool size of 40.
• --max-genotype-count: maximum number of genotypes to consider at any site:
• 
  With P la ploidy of the sample (previous formula) and A () the allele count.

4.3.1 SNPs filtering

1. Crete a subset keeping only SNPs:

gatk SelectVariants \
    -R reference_genome.fasta \
    -V input.vcf \
    -O output.vcf \
    --select-type-to-include SNP

• -R: genome reference;
• -V: .vcf input file;
• -O: output file;
• --select-type-to-include select a type of variant.

2. Generate a table that can be used to find filter thresholds:

	gatk VariantsToTable \
		-R reference_genome.fasta \
		-V input.vcf \
		-O output.table \
		# scores to save to the table
		-F CHROM \
		-F POS \
		-F QUAL \
		-F AC \
		-F AF \
		-F DP \
		-F QD \
		-F MQ \
		-F FS \
		-F SOR \
		-F MQRankSum \
		-F ReadPosRankSum

• -R: reference;
• -V: .vcf input file;
• -O: output file;
• -F: selected field.

The best way to choose the right filter, is to plot the several scores across the VCF file, for example using ggplot2 library in R. You can find here the plot used to define some scores:

3. Use filter on vcf files:

	gatk VariantFiltration \
		-R reference_genome.fasta \
		-V input.vcf \
		-O output.vcf \
		# following results observed in plots
		-filter "QUAL < 70" --filter-name Low_Qual\
		-filter "DP < 90" --filter-name Low_Cov \
		-filter "QD < 20.0 || MQ < 28.0 || FS > 1.0 || \
		SOR > 1.4 --filter-name Secondary_filter

• -R: reference;
• -V: vcf input file;
• -O: output file;
• -filter: define a filter;
  • QUAL: Phred-scaled quality score for the assertion made in alternative;
  • DP: read depth at this position for this sample;
  • QD: this is the variant confidence (from the QUAL field) divided by the unfiltered depth of non-hom-ref samples;
  • MQ: mapping quality;
  • FS: this is the Phred-scaled probability that there is strand bias at the site;
  • SOR: this is another way to estimate strand bias using a test similar to the symmetric odds ratio test;
• --filter-name: used to assign a name to the filter.

4. Extract SNPs who passed the filters:

gatk SelectVariants \
		-R reference_genome.fasta \
		-V input.vcf \
		-O output.vcf \
		--exclude-filtered

• -R: reference;
• -V: vcf input file;
• -O: output file;
• --exclude-filtered: exclude SNPs marked to be filtered.

4.3.2 Get SNPs that are in transcripts (exons)

1. From the .gff3 file of the reference, we can extract a subset having only exons:

tail -n +2 Drosophila-suzukii-annotation-3-ws.gff3 | \
awk -F "\t" '{if($3=="exon") print $0}' > ~/exon.gff3

• tail -n +2: to skip gff3 header;
• awk: if the third field has 'exon' tag, it print the line in the exon.gff3 file.

2. Convert the .gff3 into .bed format:

gff2bed < exon.gff3 > exon.bed

3. Intersect the .vcf file with the .bed file:

intersectBed -a input.vcf -b exon.bed -header > output.vcf

• -a: .vcf input file;
• -b: .bed input file;
• -header: if the .bed file has header.

4. Count the number of gene for a given sample:

• for DNA data we have used gene_count.py that can be founded in the scripts/ folder;
• for the RNA data we have used the following command (car genes are used as chromosome/contig in these files):

grep -v "#" file.vcf | cut -f1 | sort | uniq | wc -l

4.3.3 Get SNPs that are in DE genes

List of the DE genes can be found here:

Location: pedago-ngs

/localdata/pandata/students/M2_projet_15/DE_gene

This list is obtained by the M. Tabourin's work.

1. Obtain a .gff3 file containing only DE genes:

python3 get_de.py

• get_de.py: it is a script written by us. It allows, from the gff3 and the list.txt having DE genes, to obtain a .gff3 file having only DE genes.

2. Then we have transformed the obtained .gff3 file into .bed file using the command 3 at the previous paragraph and we have used it to obtain SNPs in DE genes using command 4 of the previous paragraph.

3. Finally we have compute the intersect of the vcf files from DNA and RNA. Unfortunately, these vcf files (DNA and RNA) were generated with a different reference, so we have wrote a script to convert genomic positions of the RNA vcf files to DNA positions. It can be found in the scripts/ folder as convert_vcf_pos.py. Then we have run the following commands:

# Compress VCF files
bgzip input_DNA_file.vcf
bgzip input_RNA_file.vcf

# Build indexes for compressed files
tabix -p vcf input_DNA_file.vcf.gz
tabix -p vcf input_RNA_file.vcf.gz

# run the intersection between the two files
bcftools isec -c snps -p output/ input_DNA_file.vcf.gz input_RNA_file.vcf.gz

4.3.4 Results

Location of .vcf result files: pedago-ngs

# DNA
/localdata/pandata/students/M2_projet_15/GATK_pipeline/gatk_ploidy 
# RNA
/localdata/pandata/students/M2_projet_15/RNAseq/vcf 

4.3.4.1 SNPs founded on exons

	G0	cerise	fraise	gene count
DNA	59,055	50,077	52,243	9,837
RNA	28,150	29,823	31,052	5,248

In the DNA data we find almost double compared to the RNA, but this can be easily explained by the number of genes that is also double. Nevertheless we observe a decrease from G0 to G12 for DNA, which is rather consistent with the basic hypothesis of having a selection effect. On the other hand, in RNA, the number does not decrease.

4.3.4.1 SNPs founded on exons of DE genes

	G0	cerise	fraise
DNA	536	415	492
RNA	17	36	51
intersect	1	8	7

On differentially expressed genes, there are many more SNPs on the DNA than on the RNA. It is suspected that this may be due to having poolseq data for DNA. We find again a decrease between G0 and G12 for DNA (is it significant?) but not between G0 and G7 for RNA.

The SNP founde both in RNA and DNA is on Contig 9. The 8 SNPs for the cherry are founded on Contig 9 (2), Contig 44 (5) and Contig 49 (1). The 7 SNPs for the strawberry are founded on Andromeda (1), Cepheus (1), Contig (5), Contig 9 (1), Contig 10 (1) and Contig 44 (2).

5. Diversity & Differentiation Analysis

Installating PoPoolation: Diversity Analysis

All the steps come from PoPoolation website (https://sourceforge.net/p/popoolation/wiki/browse_pages/) and Dr. Kofler presentation (https://www.kofler.or.at/bioinformatic/wp-content/uploads/2018/07/pooledAnalysis_part1.pdf)

svn checkout https://svn.code.sf.net/p/popoolation/code/trunk popoolation-code
cd popoolation-code
svn update

1. Remove low quality alignments

time samtools view -q 20 -f 0x0002 -F 0x0004 -F 0x0008 -b G0_MarkDuplicated.bam > G0_q20_MarkDuplicated.bam #26m

• -q 20 only keep reads with a mapping quality higher than 20 (remove ambiguously aligned reads)
• -f 0x0002 only keep proper pairs (remember flags from sam file)
• -F 0x0004 remove reads that are not mapped
• -F 0x0008 remove reads with an un-mapped mate

Note -f means only keep reads having the given flag and -F discard all reads having the given flag.

2. Creating a mpileup file

time samtools mpileup -B -Q 0 -f Drosophila-suzukii-contig.fasta G0_q20_MarkDuplicated.bam > G0.mpileup #20m

• -B disable BAQ computation (base alignment quality)
• -Q skip bases with base quality smaller than the given value
• -f path to reference sequence

MPILEUP FILE:

• col1 reference chromosome
• col2 position
• col3 reference character
• col4 coverage
• col5 bases for the given position (’.’ identical to reference character -forward strand; ’,’ identical to reference - reverse strand)
• col6 quality for the bases

3. Filtering indels

time perl ~/Documents/Projet_S3/Popoolation/popoolation-code/basic-pipeline/identify-genomic-indel-regions.pl --indel-window 5 --min-count 2 --input G0.mpileup --output indels_G0.gtf #25m

• –indel-window how many bases surrounding indels should be ignored
• –min-count minimum count for calling an indel. Note that indels may be sequencing errors as well

time perl ~/Documents/Projet_S3/Popoolation/popoolation-code/basic-pipeline/filter-pileup-by-gtf.pl --input G0.mpileup --gtf indels_G0.gtf --output G0.idf.mpileup #20m

4. Subsampling to uniform coverage

Several population genetic estimators are sensitive to sequencing errors. For example a very low Tajima’s D, usually indicative of a selective sweep, may be, as an artifact, frequently be found in highly covered regions because these regions have just more sequencing errors. To avoid these kinds of biases we recommend to subsample to an uniform coverage.

time perl ~/Documents/Projet_S3/Popoolation/popoolation-code/basic-pipeline/subsample-pileup.pl --min-qual 20 --method withoutreplace --max-coverage 50 --fastq-type sanger --target-coverage 10 --input G0.idf.mpileup --output G0.ss10.idf.mpileup #343m

• –min-qual minimum base quality
• –method method for subsampling, we recommend without replacement
• –target-coverage which coverage should the resulting mpileup file have
• –max-coverage the maximum allowed coverage, regions having higher coverages will be ignored (they may be copy number variations and lead to wrong SNPs)
• –fastq-type (sanger means offset 33)

5. Calculating Tajima’s Pi

time perl ~/Documents/Projet_S3/Popoolation/popoolation-code/Variance-sliding.pl --fastq-type sanger --measure pi --input G0.ss10.idf.mpileup --min-count 2 --min-coverage 4 --max-coverage 10 --min-covered-fraction 0.5 --pool-size 80 --window-size 1000 --step-size 1000 --output G0.pi --snp-output G0.snps

• –min-coverage –max-coverage: for subsampled files not important; should contain target coverage, i.e.: 10
• –min-covered-fraction minimum percentage of sites having sufficient coverage in the given window
• –min-count minimum occurrence of allele for calling a SNP
• –measure which population genetics measure should be computed (pi/theta/D)
• –pool-size number of chromosomes (thus number of diploids times two)
• –region compute the measure only for a small region; default is the whole genome
• –output a file containing the measure (π) for the windows
• –snp-output a file containing for every window the SNPs that have been used for computing the measure (e.g. π)
• –window-size –step-size control behavior of sliding window; if step size is smaller than window size than the windows will be overlapping.

Output G0.pi:

• col 1: reference chromosome
• col 2: position of window (mean value)
• col 3: number of SNPs in the given window
• col 4: fraction of sites in the window having sufficient coverage (min ≤ x ≤ max)
• col 5: measure for the window (π)

Output G0.snps:

• col 1: reference chromosome
• col 2: position of SNP
• col 3: reference character
• col 4: coverage
• col 5-9: counts of A, T, C, G, N respectively

RESULTS:

/localdata/pandata/students/M2_projet_15/GATK_pipeline/PoPoolation

Theta Pi	G0	G12 cranberry	G12 cerise	G12 fraise
moyenne	0.00830	0.00771	0.00939	0.00864
écart type	0.01250	0.01289	0.01468	0.01418

Installating PoPoolation2: Differentiation Analysis

All the steps comes from PoPoolation website (https://sourceforge.net/p/popoolation2/wiki/Home/) and Dr. Kofler presentation (https://www.kofler.or.at/bioinformatic/wp-content/uploads/2018/07/pooledAnalysis_part2.pdf)

svn checkout https://svn.code.sf.net/p/popoolation2/code/trunk popoolation2-code
cd popoolation2-code/
svn update

1. Create a mpileup

time samtools mpileup -B -Q 0 -f Drosophila-suzukii-contig.fasta G0_q20_MarkDuplicated.bam G12_cerise_q20_MarkDuplicated.bam G12_cran_q20_MarkDuplicated.bam G12_fraise_q20_MarkDuplicated.bam > p1-4.mpileup # 77m

MPILEUP:

Very similar to the mpileup with one sample. In this multi-pileup (mpileup) three additional columns are created for each additional sample. Thus for each sample the following information is provided:

• the coverage, columns 4 + n ∗ 3
• the bases, columns 5 + n ∗ 3
• the corresponding base quality, columns 6 + n ∗ 3

2. Conversion to sync-file

In order to use PoPoolation2 we have to convert the mpileup to a sync file. This may seem unnecessary to you, but it serves to speed up the analysis, because the time consuming part of the analysis - parsing of the mpileup file - needs just to be performed once.

time java -jar ~/Documents/Projet_S3/Popoolation/popoolation2-code/mpileup2sync.jar --input p1-4.mpileup --output p1-4.sync --fastq-type sanger --min-qual 20 --threads 15 #78m

THE SYNC-FILE:

• col 1: reference chromosome

• col 2: position

• col 3: reference character

• col 4: allele counts for first population

• col 5: allele counts for second population

• col n: allele counts for n-3 population

• Allele counts are in the form ”A:T:C:G:N:del”

• the sync-file provides a convenient summary of the allele counts of several populations (there is no upper threshold of the population number).

• subsampling to an uniform coverage with the PoPoolation2 pipeline should be done using such a sync-file

3. Calculating the Fst

time perl ~/Documents/Projet_S3/Popoolation/popoolation2-code/fst-sliding.pl --window-size 1 --step-size 1 --suppress-noninformative --input p1-4.sync --min-covered-fraction 1.0 --min-coverage 4 --max-coverage 120 --min-count 3 --output fst_tot.txt --pool-size 80 #386m

• --suppress-noninformative suppresses output for windows not containing a SNP. When applied to windows of size one, this option suppresses output for bases that are no SNP.

OUTPUT OF THE Fst SCRIPT:

• col 1: reference chromosome
• col 2: position
• col 3: window size (1 for single SNPs)
• col 4: covered fraction (relevant for minimum covered fraction)
• col 5: average minimum coverage for the window across all populations (the higher the more reliable the estimate)
• col 6: pairwise Fst comparing population 1 with population 2
• col 7: etc for ALL pairwise comparisons of the populations present in the sync file

RESULTS:

We separate the fst_tot.txt file in 3 files each containing one comparison with G0.

cat fst_tot.txt | sed 's/1:2=//' | awk  -F" " '{print $1" " $2" " $6}'> Fst_1-2.txt
cat fst_tot.txt | sed 's/1:3=//' | awk  -F" " '{print $1" " $2" " $7}'> Fst_1-3.txt
cat fst_tot.txt | sed 's/1:4=//' | awk  -F" " '{print $1" " $2" " $8}'> Fst_1-4.txt

/localdata/pandata/students/M2_projet_15/GATK_pipeline/PoPoolation2

Fst_1-2.txt: G0-G12_cerise

Fst_1-3.txt: G0-G12_cran

Fst_1-4.txt: G0-G12_fraise

Fst totalité des SNP	G0-G12 cranberry	G0-G12 cerise	G0-G12 fraise
moyenne	0.02347	0.03759	0.03145
écart type	0.03630	0.05538	0.04696

Fst gènes DE	G0-G12 cranberry	G0-G12 cerise	G0-G12 fraise
moyenne	0.02188	0.03429	0.02229
écart type	0.03374	0.04879	0.03373

4. Calculating Fst for DE genes:

We write a script using Fst results and a list of DE gene in a bed format to calculate Fst for DE gene. As we use dichotomous search, the Fst file must be sorted like the pipeline does. You find it on the github repo:

 scripts/Fst_DE_Genes.py

5. Visualisation of Fst on chromosomes:

We write a function in R to visualize the Fst on a chosen chromosome. You find it on the github repo:

 scripts/VisualisationFst.R

6. TE Analysis

All the steps are included in a tool named dnaPipeTE

• Uniform samplings of the reads to produce low coverage data sets (read sampling to have <1X coverage of the genome to keep only repeated regions)
• Trinity, repeated assembly of contigs
• RepeatMasker, contigs annotation
• Blastn, repeat quantification

Figure from the tool article

6.1 Running DnaPipeTE

#dnaPipeTE: version 1.3 (uses Perl 5, R v3.0.2, Python v3.8.5, Trinity v2.5.1, 
			      RepeatMasker v4.0.5 including RMblastn, ncbi-blast v2.2.28+)

We had a lot of dependancies problems running dnaPipeTE so we used the docker version.

1. Copy input data

Two different types of input are necessary to run the pipeline :

• The fastq files for each samples
• Drosophila_Transposable_Element_all.fasta which is the Drosophila TE library from Repbase

Location : pedago-ngs

/localdata/pandata/students/M2_projet_15/TE/

They all had to be in a ~/Project folder which is the communication folder between the computer and the container.

2. Starting the container

sudo docker pull clemgoub/dnapipete:latest
sudo docker run -it -v ~/Project:/mnt clemgoub/dnapipete:latest

3. Run dnaPipeTE

cd /opt/dnaPipeTE
python3 dnaPipeTE.py \ 
		-input /mnt/data/sample_1.fastq.gz 
		-output /mnt/output_dnapipete/sample 
		-cpu 6 
		-genome_size 270000000 
		-genome_coverage 0.2 
		-sample_number 2 
		-RM_lib /mnt/Drosophila_Transposable_Element_all.fasta

• -input: input fastq or fastq.gz files (single end only);
• -output: complete path with name for the outputs;
• -cpu: maximum number of cpu to use;
• -genome_size: here we know that D.suzukii is 270Mb long;
• -genome_coverage: for this genome size 0.2X is recommended in the Supplementary Material of the tool article;
• -sample_number: number of Trinity iterations (2 is the minimum and recommended number of iterations);
• -RM_lib: path to custom repeat library for RepeatMasker.

6.2 Results

Example of sampling for G0 :

total number of reads: 121247626
maximum number of reads to sample:  710526
fastq :  /mnt/G0-MTP_1.fastq
sampling 2 samples of max 355263 reads to reach coverage...
53999976 bases sampled in 355263 reads
s_G0-MTP_1.fastq done.
53999976 bases sampled in 710526 reads
s_G0-MTP_1.fastq done.

The pipeline gives several useful outputs such as the Landscapes (TE divergence to estimate the insertions) and PieChart (TE families proportions in the genome). Other graphs have been created with the script graphs.R to compare TE throught different sample generations.

Other graphs can be found in the graphs.pdf document.

7. Authors

The project was developped by:

• Chloé AUJOULAT
• Tommaso BARBERIS
• Bertrand HUGUENIN-BIZOT
• Marie VERNERET

8. References

• Anand, Santosh, Eleonora Mangano, Nadia Barizzone, Roberta Bordoni, Melissa Sorosina, Ferdinando Clarelli, Lucia Corrado, Filippo Martinelli Boneschi, Sandra D’Alfonso, et Gianluca De Bellis. « Next Generation Sequencing of Pooled Samples: Guideline for Variants’ Filtering ». Scientific Reports 6, no 1 (27 septembre 2016): 33735. https://doi.org/10.1038/srep33735.
• Chiu, Joanna C, Xuanting Jiang, Li Zhao, Christopher A Hamm, Julie M Cridland, Perot Saelao, Kelly A Hamby, et al. « Genome of Drosophila suzukii, the Spotted Wing Drosophila ». G3 Genes|Genomes|Genetics 3, no 12 (1 décembre 2013): 2257‑71. https://doi.org/10.1534/g3.113.008185.
• Kofler, Robert, Daniel Gómez-Sánchez, et Christian Schlötterer. « PoPoolationTE2: Comparative Population Genomics of Transposable Elements Using Pool-Seq ». Molecular Biology and Evolution 33, no 10 (octobre 2016): 2759‑64. https://doi.org/10.1093/molbev/msw137.
• Olazcuaga, Laure, Julien Foucaud, Mathieu Gautier, Candice Deschamps, Anne Loiseau, Nicolas Leménager, Benoit Facon, et al. « Adaptation and Correlated Fitness Responses over Two Time Scales in Drosophila Suzukii Populations Evolving in Different Environments ». Journal of Evolutionary Biology 34, no 8 (2021): 1225‑40. https://doi.org/10.1111/jeb.13878.
• Clément Goubert, Laurent Modolo, Cristina Vieira, Claire ValienteMoro, Patrick Mavingui, Matthieu Boulesteix, De Novo Assembly and Annotation of the Asian Tiger Mosquito (Aedes albopictus) Repeatome with dnaPipeTE from Raw Genomic Reads and Comparative Analysis with the Yellow Fever Mosquito (Aedes aegypti), Genome Biology and Evolution, Volume 7, Issue 4, April 2015, Pages 1192–1205, https://doi.org/10.1093/gbe/evv050
• Marie Tabourin et al. “Analyse de l’expression des gènes et des éléments transposableschez Drosophila suzukii suite à une expérience d’évolution expérimentale.
• Simon Andrews et al.FastQC. Babraham Institute. Babraham, UK, Jan. 2012, https://www.bioinformatics.babraham.ac.uk/projects/fastqc/.
• MultiQC: Summarize analysis results for multiple tools and samples in a single report, Philip Ewels, Måns Magnusson, Sverker, Lundin and Max Käller, Bioinformatics (2016), doi: 10.1093/bioinformatics/btw354, PMID: 27312411.
• Li H. and Durbin R. (2009) Fast and accurate short read alignment with Burrows-Wheeler Transform. Bioinformatics, 25:1754-60. [PMID: 19451168].
• Heng Li, Bob Handsaker, Alec Wysoker, Tim Fennell, Jue Ruan, Nils Homer, Gabor Marth, Goncalo Abecasis, Richard Durbin, 1000 Genome Project Data Processing Subgroup, The Sequence Alignment/Map format and SAMtools, Bioinformatics, Volume 25, Issue 16, 15 August 2009, Pages 2078–2079, https://doi.org/10.1093/bioinformatics/btp352.
• Van der Auwera GA & O'Connor BD. (2020). Genomics in the Cloud: Using Docker, GATK, and WDL in Terra (1st Edition). O'Reilly Media.
• Wickham H (2016). ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. ISBN 978-3-319-24277-4, https://ggplot2.tidyverse.org.
• R Core Team (2021). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.
• Bedops.readthedocs.io. 2021. 6.3.3.5. gff2bed — BEDOPS v2.4.40. (online) Available at: https://bedops.readthedocs.io/en/latest/content/reference/file-management/conversion/gff2bed.html (Accessed 2 December 2021).
• Bedtools.readthedocs.io. 2021. intersect — bedtools 2.30.0 documentation. (online) Available at: https://bedtools.readthedocs.io/en/latest/content/tools/intersect.html (Accessed 2 December 2021).