Workflow modified by Anonymous

Anonymous — Wed, 18 Mar 2015 14:54:16 -0000

Introduction
Requirements
Workflow
- Preparing the reference sequence
- Obtaining a TE hierarchy
- Mapping the sequences
- Identify TE insertions
- Estimate population frequencies

Introduction

A simple and cheap approach to genome-wide assess TE insertion frequencies in populations will allow addressing fundamental questions about TE dynamics and species evolution in more general.

Here we introduce a novel and cost efficient approach to estimate TE population frequencies for TE insertions that are present in the reference sequence as well as for novel TE insertions. Our approach solely requires sequencing of a single paired-end (PE) sample of a pooled population per investigated population. Our approach has the additional advantage that sequencing of pooled populations also yields genome-wide estimates of SNP frequencies and thus standard population genetics measures such as Tajima’s D can easily be calculated http://code.google.com/p/popoolation/.

Requirements

paired-end sequences of a pooled population
a reference genome
a transposable element sequence database (i.e.: a fasta file containing sequences of TEs)
Perl
BWA
SAMtools
RepeatMasker
PoPoolation TE

Workflow

Preparing the reference sequence

Download a reference genome. Remove unnecessary comments from the fasta headers

cat dmel-all-chromosome-r5.31|awk '{print $1}' > ../wg/dmel-5.31-short.fasta

Download the transposon sequences. Remove unnecessary comments from the fasta header
Filter TE elements that are to short (e.g.: < 40bp)

RepeatMask the reference genome using the transposon sequences as a custom library

perl RepeatMasker -no_is -nolow -norna --lib te-sequences.fasta -pa 4 reference-genome.fasta

Merge the repeat-masked reference genome and the TE sequences

cat reference-genome.fasta.masked te-sequences.fasta > combined-reference.fasta

Obtaining a TE hierarchy

Create a TE hierarchy containing one entry for EVERY sequence in the TE sequences fasta-file! The depth of the hierarchy, the classification of the elements and the names for different hierarchy-levels is up to the user!

example of an hierarchy file:

insert  id  family  superfamily suborder    order   class   problem
FBti0015567 Tirant  Tirant  Tirant  LTR LTR RNA 0   
FBti0018861 17.6    17.6    17.6    LTR LTR RNA 0   
FBti0018862 17.6    17.6    17.6    LTR LTR RNA 0   
FBti0018865 297 297 297 LTR LTR RNA 0   
FBti0018869 3S18    3S18    3S18    LTR LTR RNA 0
FBti0018957 baggins baggins baggins non-LTR non-LTR RNA 0
FBti0018944 Tc1 Tc1 Tc1 TIR TIR DNA 0

The first row specifies the names of the hierarchy-levels in ascending order. The problem column may be ignored All other columns refer to individual TE sequences and their hierarchical status. For every entry in the TE-sequences fasta file the TE-hierarchy has to contain exactly one entry. Note that many individual entries may be present for TE insertions of the same family. The software tries to handle the resulting sequence ambiguity by moving in this TE-hierarchy.

Mapping the sequences

Index the reference genome
```
bwa index combined-reference.fasta
```

Map the PE reads separately using BWA SW (Smith-Waterman algorithm)

bwa bwasw -t 6 combined-reference.fasta ../reads/Jul_s_5_1_sequence.txt > Jul_s_5_1_seq.sam
bwa bwasw -t 6 combined-reference.fasta ../reads/Jul_s_5_2_sequence.txt > Jul_s_5_2_seq.sam

Create paired-end information using samro (PoPoolation TE)

perl samro.pl --sam1 Jul_s_5_1_seq.sam --sam2 Jul_s_5_2_seq.sam --fq1 ../reads/Jul_s_5_1_s
equence.txt --fq2 ../reads/Jul_s_5_2_sequence.txt --output pe-reads.sam

Sort the sam file

The sam-file needs to be sorted with samtools

samtools view -Sb pe-reads.sam | samtools sort - pe-reads.sorted
samtools view pe-reads.sorted.bam > pe-reads.sorted.sam

Identify TE insertions

Identify forward and reverse insertions

perl identify-te-insertsites.pl --input pe-reads.sorted.sam --te-hierarchy-file te-hierarchy.txt --te-hierarchy-level family --narrow-range 75 --min-count 3 --min-map-qual 15 --output te-fwd-rev.txt

Identify the positions of poly-N stretches in the repeat-masked reference genome. PoPoolation TE calculates the distance between forward and reverse insertions and ignores poly-N stretches for calculating these distances.
```
perl genomic-N-2gtf.pl --input combined-reference.fasta > poly_n.gtf
```

Crosslink forward and reverse insertions thus obtaining TE insertions

perl crosslink-te-sites.pl --directional-insertions te-fwd-rev.txt --min-dist 74 --max-dist 250 --output te-inserts.txt --single-site-shift 100 --poly-n poly_n.gtf --te-hierarchy te-hierarchy.txt --te-hier-level order

Optional: Use the known TE insertions to improve the crosslinking of forward and reverse insertions

perl update-teinserts-with-knowntes.pl --known known-te-insertions.txt --output te-insertions.updated.txt --te-hierarchy-file te-hierarchy.txt --te-hierarchy-level family --max-dist 300 --te-insertions te-inserts.txt --single-site-shift 100

The known TE insertion file has to look like in the following example:

2L      F       20740303        FBti0015567
2L      R       20748120        FBti0015567
2R      F       5614174 FBti0018861
2R      R       5621667 FBti0018861

Column1 is the reference chromosome, column2 the direction of the insertion either forward (F) or reverse (R), column3 is the position and column4 is the fasta-header of the corresponding TE sequence (here the FlyBase ID)

Estimate population frequencies

perl estimate-polymorphism.pl --sam-file pe-reads.sorted.sam --te-insert-file te-insertions.updated.txt --te-hierarchy-file te-hierarchy.txt --te-hierarchy-level family --min-map-qual 15 --output te-polymorphism.txt

Optional: Filter the file

perl filter-teinserts.pl --te-insertions te-polymorphism.txt --output te-poly-filtered.txt --discard-overlapping --min-count 10

Voila: the output file

4       135417  F       G4      0.75    non-LTR -       -       135302  135317  0.75    4       3       1       1       -       -       -       -       -       -       -
4       135906  FR      G5      1       non-LTR FBti0020406     ncorrdist=105   135348  135447  1       17      17      0       0       136365  136464  1       74      74      0       0
4       135333  R       gypsy12 0.0784313725490196      LTR     -       -       -       -       -       -       -       -       -       135433  135491  0.0784313725490196      51      4       47      0
4       145298  FR      HB      1       TIR     FBti0062857     ncorrdist=99    144489  144588  1       23      23      0       0       146008  146085  1       47      47      0       1
4       146710  FR      Tc1     1       TIR     FBti0019474     ncorrdist=89    145963  146062  1       23      23      0       0       147358  147390  1       14      14      0       0

Col1: the reference sequence ID
Col2: position in the reference sequence
Col3: is the TE insertion supported by a forward (F), by a reverse (R) or by both (FR) insertions
Col4: family of the TE insertion
Col5: population frequency (1..fixed)
Col6: order of the TE insertion
Col7: ID if the TE insertion is present in the reference genome (e.g.: FlyBase ID)
Col8: comment
Col9: start of the range of the forward insertion
Col10: end of the range of the forward insertion
Col11: population frequency estimated by the forward insertion
Col12: coverage of the forward insertion
Col13: TE-presence reads of the forward insertion
Col14: TE-absence reads of the reverse insertion
Col15: is the range of the forward insertion overlapping with a forward-range of another TE insertion (0..no; 1..yes)
Col16: start of the range of the reverse insertion
Col17: end of the range of the reverse insertion
Col18: coverage of the reverse insertion
Col19: TE-presence reads of the reverse insertion
Col20: TE-absence reads of the reverse insertion
Col21: is the range of the reverse insertion overlapping with the reverese-range of another TE insertion (0..no; 1..yes)

Recent changes to Workflow