Session 2: Analysis of non-coding sequences

Estación Experimental de Aula Dei-CSIC, Zaragoza, Spain

1 Introduction

In this session we will continue annotating sequences, but now with a focus on non-coding sequences. If we revisit the next figure, I hope you’ll see that non-coding sequences are usually the largest fraction of genomes (Mäkinen et al. 2015):

For instance, the genome of the model grass Brachypodium distachyon is about 270Mbp long, but its genes take only about 122Mbp. The difference increases as the genomes grow larger, as you can see in the next table:

annot.stats <- read.csv(file="test_data/Ensembl_stats.tsv", sep="\t", comment.char=";")
kable(annot.stats,format.args = list(big.mark=","))

genome	genome.size	gene.space	percentage
Arabidopsis thaliana	119,667,750	77,640,474	64.9
Arabidopsis halleri	196,243,198	67,469,378	34.4
Prunus dulcis	227,498,357	94,209,302	41.4
Brachypodium distachyon	271,163,419	122,468,185	45.2
Brassica rapa	283,822,783	83,199,064	29.3
Trifolium pratense	304,842,038	133,448,980	43.8
Arabis alpina	308,032,609	45,780,039	14.9
Cucumis melo	357,857,370	99,736,823	27.9
Citrullus lanatus	365,450,462	81,219,268	22.2
Oryza sativa	375,049,285	130,066,257	34.7
Setaria viridis	395,731,502	123,250,596	31.1
Vitis vinifera	486,265,422	153,535,016	31.6
Rosa chinensis	515,588,973	117,296,819	22.8
Camelina sativa	641,356,059	214,928,593	33.5
Malus domestica	702,961,352	145,783,855	20.7
Olea europaea	1,140,987,834	153,461,985	13.4
Zea mays	2,135,083,061	168,014,276	7.9
Helianthus annuus	3,027,844,945	199,817,746	6.6
Aegilops tauschii	4,224,915,394	348,412,472	8.2
Triticum turgidum	10,463,058,104	475,613,274	4.5

Therefore, when dealing with genome variation, polymorphisms have a greater chance to occur in non-coding regions. Among these, probably the most interesting are regulatory sequences and repeated elements.

The goal of this session is to learn how regulatory sequences can be discovered in promoter sequences using statistical tests and aligned to build DNA motifs.

2 Genomic repeated sequences

The annotation of Transposable Elements (TEs) within plant genomes can help in the interpretation of observed phenotypes, as sometimes TEs affect the expression of neighbor genes, and in computational tasks such as promoter whole genome alignment, promoter or pan-genome analyses.

Usually TEs are annotated by alignment to curated libraries of repeated elements such as RepetDB (Amselem et al. 2019), where each sequence or element is classified according to the Wicker classification (Wicker et al. 2007). Class I elements are “copy and paste”, while Class II are “cut and paste”. The next figure summarizes this taxonomy of TEs, which resembles that of protein domains:

More recently, there are other approaches which do not require any curation; instead, these simply identify repeated elements by counting words (\(K\)-mers) along the genome. An example of such tools is the Repeat Detector (Girgis 2015), which has been used to routinely mask plant genomes (Contreras-Moreira et al. 2021).

The next table summarizes the fraction of repeated sequences in diverse plant genomes, as annotated with REdat, Red and the original papers describing those genomes:

annot.stats <- read.csv(file="test_data/Ensembl_repeats.tsv", sep="\t", comment.char=";")
annot.stats = annot.stats[,c(1,2,6,8,10)]
kable(annot.stats,format.args = list(big.mark=","))

genome	size	perc_REdat	perc_Red	perc_literature
Arabidopsis thaliana	119,667,750	14.2	36.7	19.0
Arabidopsis halleri	196,243,198	15.5	31.1	32.7
Prunus dulcis	227,498,357	6.5	33.4	37.6
Brachypodium distachyon	271,163,419	27.4	31.1	21.4
Brassica rapa	283,822,783	8.6	32.8	32.3
Trifolium pratense	304,842,038	11.0	30.0	41.8
Arabis alpina	308,032,609	15.1	37.7	47.9
Cucumis melo	357,857,370	8.2	39.9	44.0
Citrullus lanatus	365,450,462	6.8	40.8	45.2
Oryza sativa	375,049,285	32.3	37.0	35.0
Setaria viridis	395,731,502	18.5	40.8	46.0
Vitis vinifera	486,265,422	9.0	40.0	41.4
Rosa chinensis	515,588,973	8.4	48.1	67.9
Camelina sativa	641,356,059	15.8	36.0	28.0
Malus domestica	702,961,352	9.1	41.6	59.5
Olea europaea	1,140,989,389	18.0	45.2	43.0
Zea mays	2,135,083,061	59.5	79.0	85.0
Helianthus annuus	3,027,844,945	10.0	73.6	74.7
Aegilops tauschii	4,224,915,394	68.7	81.8	85.9
Triticum turgidum	10,463,058,104	71.7	82.2	82.2

When the repeats themselves are not of interest for subsequent analyses, they are said to be masked out. This means that they are marked so that they can be avoided. Hard-masking means replacing the sequences of repeats with polyN oligonucleotides. Soft-masking means leaving the repeated sequences in lower-case:

ACTAGACTACGNNNNNNNNNNNATATATCA  # hard-masked
ACTAGACTACGtttttttttttATATATCA  # soft-masked

3 DNA-binding proteins and regulatory sequences

The regulation of gene expression is one of the fundamental topics in Genetics. Here you will learn about transcription factors (TFs) and cis-regulatory elements (CREs). TFs are proteins that bind specifically to DNA sequences called CREs and affect the expression of nearby genes.

3.1 Protein-DNA recognition

DNA-binding proteins contain DNA-binding domains and have a specific or general affinity for either single or double stranded DNA. Here we will concentrate mostly on transcription factors, which generally recognize cis-regulatory elements in double-stranded DNA molecules.

3.1.1 Dissecting a protein-DNA interface

Transcription factors recognize target DNA sequences through a binding interface, composed of protein residues and DNA stretches in intimate contact. The best descriptions of protein-DNA interfaces are provided by structural biology, usually by X-ray or NMR experiments.