Transposable element annotation on JeStream

Tools prerequiste :

scripts and intermediate files used to annotate TEs in Jiao et al. 2016

LTR Retrotransposons

scripts in ltr

Software needed:

• ncbi blast+

• genometools, (download), need to pass 64bit=yes with-hmer=yes threads=yes to make, make install for ltrdigest hmm searches in parallel. I also had to pass cairo=no as well because I didn't have the right cairo libraries and it wouldn't compile otherwise

• silix, (download), need to compile with --enable-mpi and --enable-verbose

• hmmer (genometools with download and compile hmmer2 if you run make with-hmmer=yes)

Files needed, can be downloaded by get_tRNA_hmm_dbs.sh in ltr directory:

• download hmms of TE protein coding domains from gydb in directory gydb_hmms, will be used to identify protein coding domains of TE models

  -need to fix a hmm with name ty1/copia because this is used as a filename by ltrdigest. to remove the forward slash: sed -i "s#ty1/copia#ty1-copia#g" gydb_hmms/GyDB_collection/profiles/AP_ty1copia.hmm

• download tRNAs of all eukaryotes

SINEs

Scripts in sine/

Software needed:

• SINE-Finder, download (This is a supplemental file at The Plant Cell; need to make executable, and rename to sine_finder.py)

  • I cannot make SINE-Finder function on reverse sequences. So I'm reporting SINEs only on the forward stand here, and will pick up sequences on the reverse strand with RepeatMasker.

LINEs

Scripts in line/

Software needed:

• MGEScanNonLTR I use the version generated for Galaxy here.

TIR including MITEs

Scripts in tir/

Software needed:

• detectMITE, download

• mTEA, genometools (see above, already installed for ltr annotation)

  • mTEA needs fasta36 (specifically ggsearch36), bioperl, blast, muscle, supplied blogo directories to be put into PERL5LIB and PATH

Helitrons

Scripts in helitron/

Software needed:

• HelitronScanner

Finding Homologous Fragments from Degraded TEs

Software needed:

• RepeatMasker, with prerequisites here

Step 1: Git clone the repo

 $ git clone https://github.com/mcstitzer/maize_v4_TE_annotation.git
 $ cd maize_v4_TE_annotation/
 $ ls
helitron  line  ltr  README.md  sine  tir

Step 2: To predict structural LTRs

2.1 predict LTRs:

• download tRNA and GyDb HMMs using get_tRNA_hmm_dbs.sh, which are needed forltrdigest

• but LTR TEs are nested, so we need to remove these copies and rerun. This is done in mask_subtract

$ cd ltr
$ sh get_tRNA_hmm_dbs.sh 

This will download the tRNA database for all Eukaryotes

• run_ltrharvest.sh runs ltrharvest and ltrdigest on the genome

$ cd ltr
$ sh get_tRNA_hmm_dbs.sh