RSAT::Plants updated (Dec2022)

Hi, 

if you use the Plants server of the Regulatory Sequence Analysis Tools (RSAT), you might want to know that it has just been updated. Here's a short summary of the changes:

• The updated URL is https://rsat.eead.csic.es/plants
•  It now supports HTTPS connections powered by certbot
•  It now uses the source code at https://github.com/rsa-tools/rsat-code (I have updated some documentation along the way)
•  Nine new species have been imported from Ensembl Plants: Lolium perenne, Brassica juncea, Echinochloa crusgalli, Digitaria exilis, Vigna unguiculata, Brassica rapa ro18, Corylus avellana, Ficus carica, Lactuca sativa
•  One species renamed: Physcomitrium patens
•  Three updated with a new assembly: Vitis vinifera, Triticum urartum, sunflower
• This leaves the total number of supported assemblies in 100; you can see their stats at https://rsat.eead.csic.es/plants/data/stats
• Most species now correspond to release 55 of Ensembl Plants, but note that the sequence data is unchanged in many cases. This means that, for instance, that Hordeum_vulgare.MorexV3_pseudomolecules_assembly.52 becomes        Hordeum_vulgare.MorexV3_pseudomolecules_assembly.55, but the sequence is exactly the same.         
  

  

Have a nice break,

Bruno

Mapeando sobre un pangenoma de arroz con minigraph

Hola,

hace unos meses contruí un pangenoma a partir de 15 genomas de arroz obtenidos de Ensembl Plants. Para ello he probado el software minigraph, descrito en https://doi.org/10.1186/s13059-020-02168-z , que es una de las herramientas disponibles para construir un grafo genómico, en este caso por medio de inserciones y deleciones sobre el genoma de referencia (naranja en la figura).

 

Figura. Mapeo de lecturas sobre un grafo pangenómico, figura tomada de https://doi.org/10.1186/s13059-020-02168-z

Hoy quería resumir aquí cómo se hace por si le ayuda a alguien.

El primer paso es construir el grafo a partir de varios genomas individuales, de arroz en este caso. Para ello deberás partir de ficheros FASTA donde cada cromosoma tenga un nombre único. Eso se puede lograr por ejemplo agregando al nombre de cromosoma original el identificador o accession de cada genoma:

# 1) prepare genome FASTA files, making sure chr names are unique
mkdir fasta
while read core; do 
	echo $core; 
	perl -lne 'BEGIN{ if($ARGV[0] =~ /sativa_([^_]+)/){ $acc=$1 }} if(/^>(\S+)/){ print ">$1_$acc" } else {print}' ${acc}.fna > ${acc}.uniq.fna 
done < ../liftover/list_cores.txt

# build the graph
bsub -M 40G -n 10 -cwd soft/minigraph/minigraph -xggs -t 10 oryza_sativa_core_48_101_7.fna fasta/*.fna -o oryza_sativa.gfa

Este proceso genera la siguiente salida:

[M::main::0.702*0.84] loaded the graph from "oryza_sativa_core_48_101_7.fna"
[M::mg_index::9.913*1.50] indexed the graph
[M::mg_opt_update::10.576*1.47] occ_weight=20, occ_max1=178; 95 percentile: 2
[M::ggen_map::11.491*1.42] loaded file "fasta/oryza_sativa_Azucena_core_48_101_1.fna"
[M::ggen_map::168.948*6.05] mapped 37 sequence(s) to the graph
[M::mg_ggsimple::170.379*6.01] inserted 15028 events, including 39 inversions
[M::mg_index::180.913*5.75] indexed the graph
...
[M::main] Real time: 13152.427 sec; CPU: 82007.142 sec; Peak RSS: 47.687 GB

Ahora podemos probar a mapear secuencias de cDNA sobre el grafo:

# read
https://twitter.com/zhigui_bao/status/1417028758725222400
https://github.com/lh3/minigraph/issues/37

# Note in rice -N 0 /-N 100 made no difference!
soft/minigraph/minigraph -t 4 -j 0.02 oryza_sativa.gfa -N 100 \
	oryza_nivara.cdna.fna | sort -k1,1 -k10,10nr > Onivara.cdna.graph.sort.gaf
soft/minigraph/minigraph -t 4 -j 0.02 oryza_sativa.gfa -N 100 \
	oryza_sativa.cdna.fna | sort -k1,1 -k10,10nr > Osativa.cdna.graph.sort.gaf
soft/minigraph/minigraph -t 4 -j 0.02 oryza_sativa.gfa -N 100 \
	oryza_indica.cdna.fna | sort -k1,1 -k10,10nr > Oindica.cdna.graph.sort.gaf

Hasta pronto, Bruno