View source for Sequence analysis

{{Short description|Identification and study of genomic sequences}}
{{distinguish|sequential analysis|sequence analysis of synthetic polymers|Sequence analysis in social sciences}}
In [[bioinformatics]], '''sequence analysis''' is the process of subjecting a [[DNA sequence|DNA]], [[RNA]] or [[peptide sequence]] to any of a wide range of analytical methods to understand its features, function, structure, or evolution. It can be performed on the entire genome, transcriptome or proteome of an organism, and can also involve only selected segments or regions, like tandem repeats and transposable elements. Methodologies used include [[sequence alignment]], searches against [[biological database]]s, and others.<ref>{{Durbin 1998}}</ref>

Since the development of methods of high-throughput production of gene and protein sequences, the rate of addition of new sequences to the databases increased very rapidly. Such a collection of sequences does not, by itself, increase the scientist's understanding of the biology of organisms. However, comparing these new sequences to those with known functions is a key way of understanding the biology of an organism from which the new sequence comes. Thus, sequence analysis can be used to assign function to coding and non-coding regions in a biological sequence usually by comparing sequences and studying similarities and differences. Nowadays, there are many tools and techniques that provide the sequence comparisons (sequence alignment) and analyze the alignment product to understand its biology.Sequence analysis in [[molecular biology]] includes a very wide range of processes:{{cn|date=August 2025}}

# The comparison of sequences to find similarity, often to infer if they are related ([[Sequence homology|homologous]])
# Identification of intrinsic features of the sequence such as [[active site]]s, [[post translational modification]] sites,  [[Sequence motif|gene-structures]], [[reading frame]]s, distributions of [[intron]]s and [[exon]]s and [[Gene regulation|regulatory elements]]
# Identification of sequence differences and variations such as [[point mutation]]s and [[single nucleotide polymorphism]] (SNP) in order to get the [[genetic marker]].
# Revealing the evolution and [[genetic diversity]] of sequences and organisms
# Identification of molecular structure from sequence alone.

== History ==

Since the very first sequences of the [[insulin]] protein were characterized by [[Fred Sanger]] in 1951, biologists have been trying to use this knowledge to understand the function of molecules.<ref name="pmid14886310">{{cite journal |author1=Sanger F |author2=Tuppy H |title=The amino-acid sequence in the phenylalanyl chain of insulin. I. The identification of lower peptides from partial hydrolysates |journal=Biochem. J. |volume=49 |issue=4 |pages=463–81 |date=September 1951 |pmid=14886310 |pmc=1197535 |doi= 10.1042/bj0490463}}</ref><ref name="pmid14886311">{{cite journal |author1=SANGER F |author2=TUPPY H |title=The amino-acid sequence in the phenylalanyl chain of insulin. 2. The investigation of peptides from enzymic hydrolysates |journal=Biochem. J. |volume=49 |issue=4 |pages=481–90 |date=September 1951 |pmid=14886311 |pmc=1197536 |doi= 10.1042/bj0490481}}</ref> He and his colleagues' discoveries contributed to the successful sequencing of the first DNA-based genome.<ref>{{cite journal|last1=Sanger|first1=F|last2=Nicklen|first2=S|last3=Coulson|first3=AR|title=DNA sequencing with chain-terminating inhibitors|journal=Proc Natl Acad Sci U S A|date=December 1977 |issue=12|pages=441–448|pmid=271968|volume=74|pmc=431765|doi=10.1073/pnas.74.12.5463|bibcode=1977PNAS...74.5463S|doi-access=free}}</ref> The method used in this study, which is called the "Sanger method" or [[Sanger sequencing]], was a milestone in sequencing long strand molecules such as DNA. This method was eventually used in the [[Human Genome Project]].<ref>{{cite journal|last1=Sanger|first1=F|last2=Air|first2=GM|last3=Barrell|first3=BG|last4=Brown|first4=NL|last5=Coulson|first5=AR|last6=Fiddes|first6=CA|last7=Hutchison|first7=CA|last8=Slocombe|first8=PM|last9=Smith|first9=M|title=Nucleotide sequence of bacteriophage phi X174 DNA|journal=Nature|date= February 1977|volume=265|issue=5596|pages=687–695|pmid=870828|doi=10.1038/265687a0|bibcode=1977Natur.265..687S|s2cid=4206886}}</ref>  According to [[Michael Levitt (biophysicist)|Michael Levitt]], sequence analysis was born in the period from 1969 to 1977.<ref name="pmid11323711">{{cite journal |author=Levitt M |title=The birth of computational structural biology |journal=Nature Structural & Molecular Biology |volume=8 |issue=5 |pages=392–3 |date=May 2001 |pmid=11323711 |doi=10.1038/87545 |s2cid=6519868 }}</ref> In 1969 the analysis of sequences of [[tRNA|transfer RNAs]] was used to infer residue interactions from correlated changes in the nucleotide sequences, giving rise to a model of the tRNA [[secondary structure]].<ref name="pmid5361649">{{cite journal |author=Levitt M |title=Detailed molecular model for transfer ribonucleic acid |journal=Nature |volume=224 |issue=5221 |pages=759–63 |date=November 1969 |pmid=5361649 |doi= 10.1038/224759a0|bibcode=1969Natur.224..759L|s2cid=983981 }}</ref> In 1970, Saul B. Needleman and Christian D. Wunsch published the first [[Needleman–Wunsch algorithm|computer algorithm]] for aligning two sequences.<ref name="pmid5420325">{{cite journal |author1=Needleman SB |author2=Wunsch CD |title=A general method applicable to the search for similarities in the amino acid sequence of two proteins |journal=J. Mol. Biol. |volume=48 |issue=3 |pages=443–53 |date=March 1970 |pmid=5420325 |doi= 10.1016/0022-2836(70)90057-4}}</ref> Over this time, developments in obtaining nucleotide sequence improved greatly, leading to the publication of the first complete genome of a bacteriophage in 1977.<ref name="pmid870828">{{cite journal  |vauthors=Sanger F, Air GM, Barrell BG, etal |title=Nucleotide sequence of bacteriophage phi X174 DNA |journal=Nature |volume=265 |issue=5596 |pages=687–95 |date=February 1977 |pmid=870828 |doi= 10.1038/265687a0|bibcode = 1977Natur.265..687S |s2cid=4206886 }}</ref> Robert Holley and his team in Cornell University were believed to be the first to sequence an RNA molecule.<ref>{{cite journal|last1=Holley|first1=RW|last2=Apgar|first2=J|last3=Everett|first3=GA|last4=Madison|first4=JT|last5=Marquisee|first5=M|last6=Merrill|first6=SH|last7=Penswick|first7=JR|last8=Zamir|first8=A|title=Structure of a Ribonucleic Acid|journal=Science|date=May 1965|volume=147|issue=3664|pages=1462–1465|pmid=14263761|doi=10.1126/science.147.3664.1462|bibcode=1965Sci...147.1462H|s2cid=40989800}}</ref>

== Overview of nucleotide sequence analysis (DNA and RNA) ==

Nucleotide sequence analyses identify functional elements like protein binding sites, uncover genetic variations like SNPs, study gene expression patterns, and understand the genetic basis of traits. It helps to understand mechanisms that contribute to processes like replication and transcription. Some of the tasks involved are outlined below.{{cn|date=August 2025}}

=== Quality control and preprocessing ===
Quality control assesses the quality of sequencing reads obtained from the sequencing technology (e.g. [[Illumina, Inc.|Illumina]]). It is the first step in sequence analysis to limit wrong conclusions due to poor quality data. The tools used at this stage depend on the sequencing platform. For instance,  FastQC checks the quality of short reads (including RNA sequences), Nanoplot or PycoQC are used for [[third-generation sequencing|long read sequences]] (e.g. Nanopore sequence reads), and MultiQC aggregates the result of FastQC in a webpage format.<ref>{{cite web |last1=Batut |first1=Bérénice |last2=Doyle |first2=Maria |last3=Cormier |first3=Alexandre |last4=Bretaudeau |first4=Anthony |last5=Leroi |first5=Laura |last6=Corre |first6=Erwan |last7=Robin |first7=Stéphanie |last8=nil |first8=gallantries |last9=Hyde |first9=Cameron |title=Quality Control (Galaxy Training Materials) |url=https://training.galaxyproject.org/training-material/topics/sequence-analysis/tutorials/quality-control/tutorial.html |website=Galaxy Training! |date=3 November 2023 |access-date=26 April 2024}}</ref><ref name=galaxy1>{{cite journal |last1=Hiltemann |first1=Saskia |last2=Rasche |first2=Helena |last3=Gladman |first3=Simon | display-authors = 2 |title=Galaxy Training: A Powerful Framework for Teaching! |journal=PLOS Computational Biology |date=January 2023 |volume=19 |issue=1 |article-number=e1010752 |doi=10.1371/journal.pcbi.1010752 |doi-access=free |pmid=36622853 |pmc=9829167 |bibcode=2023PLSCB..19E0752H }}</ref><ref name=galaxy2>{{cite journal |last1=Batut |first1=Bérénice |last2=Hiltemann |first2=Saskia | display-authors = 1 |title=Community-Driven Data Analysis Training for Biology |journal=Cell Systems |date=2018 |volume=6 |issue=6 |pages=752–758.e1 |doi=10.1016/j.cels.2018.05.012 |pmid=29953864 |pmc=6296361 }}</ref>

Quality control provides information such as read lengths, [[GC-content|GC content]], presence of adapter sequences (for short reads), and a quality score, which is often expressed on a [[phred quality score|PHRED scale]].<ref name=sequence_analysis>{{cite book |last1=Prijibelski |first1=Andrey B. |last2=Korobeynikov |first2=Anton I. |last3=Lapidus |first3=Alla L. |chapter=Sequence Analysis |title=Encyclopedia of Bioinformatics and Computational Biology |date=September 2018 |volume=3 |pages=292–322 |doi=10.1016/B978-0-12-809633-8.20106-4 |isbn=978-0-12-811432-2 }}</ref> If adapters or other artifacts from PCR amplification are present in the reads (particularly short reads), they are removed using software such as Trimmomatic<ref>{{cite journal |last1=Bolger |first1=Anthony M. |last2=Lohse |first2=Marc |last3=Usadel |first3=Bjoern |title=Trimmomatic: a flexible trimmer for Illumina sequence data |journal=Bioinformatics |date=April 2014 |volume=30 |issue=15 |pages=2114–2120 |doi=10.1093/bioinformatics/btu170 |pmid=24695404 |pmc=4103590 }}</ref> or Cutadapt.<ref>{{cite journal |last1=Marcel |first1=Martin |title=Cutadapt removes adapter sequences from high-throughput sequencing reads |journal=EMBnet.journal |date=2011 |volume=17 |page=10 |doi=10.14806/ej.17.1.200 }}</ref>

=== Read alignment ===

At this step, sequencing reads whose quality have been improved are mapped to a [[reference genome]] using alignment tools like BWA<ref>{{cite journal |last1=Li |first1=Heng |last2=Durbin |first2=Richard |title=Fast and accurate short read alignment with Burrows–Wheeler transform |journal=Bioinformatics |date=July 2009 |volume=25 |issue=14 |pages=1754–1760 |doi=10.1093/bioinformatics/btp324 |pmid=19451168 |doi-access=free |pmc=2705234 }}</ref> for short DNA sequence reads, minimap<ref>{{cite journal |last1=Li |first1=Heng |title=Minimap and miniasm: fast mapping and de novo assembly for noisy long sequences |journal=Bioinformatics |date=March 2016 |volume=32 |issue=14 |pages=2103–2110 |doi=10.1093/bioinformatics/btw152 |pmid=27153593 |pmc=4937194 }}</ref> for long read DNA sequences, and STAR<ref>{{cite journal |last1=Dobin |first1=Alexander |last2=Davis |first2=Carrie A. |last3=Schlesinger |first3=Felix | display-authors = 2 |title=STAR: ultrafast universal RNA-seq aligner |journal=Bioinformatics |date=October 2012 |volume=29 |issue=1 |pages=15–21 |doi=10.1093/bioinformatics/bts635 |pmid=23104886 |pmc=3530905 }}</ref> for RNA sequence reads. The purpose of mapping is to find the origin of any given read based on the reference sequence. It is also important for detecting variations or [[molecular phylogenetics|phylogenetic studies]].
The output from this step, that is, the aligned reads, are stored in compatible file formats known as SAM, which contains information about the reference genome as well as individual reads. Alternatively, [[Binary Alignment Map|BAM file]] formats are preferred as they use much less desk or storage space.<ref name=sequence_analysis/>

'''Note''': This is different from sequence alignment which compares two or more whole sequences (or sequence regions) to quantify similarity or differences or to identify an unknown sequence (as discussed below).

'''The following analyses steps are peculiar to DNA sequences:'''

=== Variant calling ===

Identifying variants is a popular aspect of sequence analysis as variants often contain information of biological significance, such as explaining the mechanism of drug resistance in an infectious disease. These variants could be single nucleotide variants (SNVs), small insertions/deletions (indels), and large [[structural variation|structural variants]]. The read alignments are sorted using [[SAMtools]], after which variant callers such as GATK<ref>{{cite journal |last1=McKenna |first1=Aaron |last2=Hanna |first2=Matthew |last3=Banks |first3=Eric | display-authors = 2 |title=The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation DNA sequencing data |journal=Genome Research |date=July 2010 |volume=20 |issue=9 |pages=1297–1303 |doi=10.1101/gr.107524.110 |pmid=20644199 |pmc=2928508 }}</ref> are used to identify differences compared to the reference sequence.

The choice of variant calling tool depends heavily on the sequencing technology used, so GATK is often used when working with short reads, while long read sequences require tools like DeepVariant<ref>{{cite journal |last1=Poplin |first1=R |last2=Chang |first2=PC |last3=Alexander |first3=D |display-authors=2 |title=A universal SNP and small-indel variant caller using deep neural networks |journal=Nature Biotechnology |date=September 2018 |volume=36 |issue=10 |pages=983–987 |doi=10.1038/nbt.4235 |pmid=30247488 }}</ref> and Sniffles.<ref>{{cite journal |last1=Sedlazeck |first1=F.J. |last2=Rescheneder |first2=P |last3=Smolka |first3=M |display-authors=2 |title=Accurate detection of complex structural variations using single-molecule sequencing |journal=Nature Methods |date=April 2018 |volume=15 |issue=6 |pages=461–468 |doi=10.1038/s41592-018-0001-7 |pmid=29713083 |pmc=5990442 }}</ref> Tools may also differ based on organism (prokaryotes or eukaryotes), source of sequence data (cancer vs [[metagenomics|metagenomic]]), and variant type of interest (SNVs or structural variants). The output of variant calling is typically in [[Variant Call Format|vcf format]], and can be filtered using allele frequencies, quality scores, or other factors based on the research question at hand.<ref name=sequence_analysis/>

=== Variant annotation ===

This step adds context to the variant data using curated information from peer-reviewed papers and publicly available databases like gnomAD and [[Ensembl genome database project|Ensembl]]. Variants can be annotated with information about genomic features, functional consequences, regulatory elements, and population frequencies using tools like ANNOVAR or SnpEff,<ref>{{cite journal |last1=Cingolani |first1=P |last2=Platts |first2=A |last3=Wang |first3=L |display-authors=2 |title=A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff |journal=Fly |date=April 2012 |volume=6 |issue=2 |pages=80–92 |doi=10.4161/fly.19695 |pmid=22728672 |pmc=3679285 }}</ref> or custom scripts and pipeline. The output from this step is an annotation file in bed or txt format.<ref name=sequence_analysis/>

=== Visualization and interpretation ===

Genomic data, such as read alignments, coverage plots, and variant calls, can be visualized using [[genome browser]]s like IGV (Integrative Genomics Viewer) or UCSC Genome Browser. Interpretation of the results is done in the context of the biological question or hypothesis under investigation. The output can be a graphical representation of data in the forms of Circos plots, volcano plots, etc., or other forms of report describing the observations.<ref name=sequence_analysis/>

DNA sequence analysis could also involve statistical modeling to infer relationships and epigenetic analysis, like identifying differential methylation regions using a tool like DSS.{{cn|date=August 2025}}

'''The following steps are peculiar to RNA sequences:'''

=== Gene expression analysis ===

Mapped RNA sequences are analyzed to estimate gene expression levels using quantification tools such as HTSeq,<ref>{{cite journal |last1=Anders |first1=Simon |last2=Pyl |first2=Paul Theodore |last3=Huber |first3=Wolfgang |title=HTSeq—a Python framework to work with high-throughput sequencing data |journal=Bioinformatics |date=January 2015 |volume=31 |issue=2 |pages=166–169 |doi=10.1093/bioinformatics/btu638 |pmid=25260700 |pmc=4287950 }}</ref> and identify differentially expressed genes (DEGs) between experimental conditions using statistical methods like [[DESeq2]].<ref>{{cite journal |last1=Love |first1=M.I. |last2=Huber |first2=W. |last3=Anders |first3=S. |title=Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 |journal=Genome Biology |date=December 2014 |volume=15 |issue=12 |page=550 |doi=10.1186/s13059-014-0550-8 |doi-access=free |pmid=25516281 |pmc=4302049 }}</ref> This is carried out to compare the expression levels of genes or isoforms between or across different samples, and infer biological relevance.<ref name=sequence_analysis/>
The output of gene expression analysis is typically a table with values representing the expression levels of gene IDs or names in rows and samples in the columns as well as standard errors and p-values. The results in the table can be further visualized using volcano plots and heatmaps, where colors represent the estimated expression level. Packages like ggplot2 in R and Matplotlib in Python are often used to create the visuals. The table can also be annotated using a reference annotation file, usually in [[gene transfer format|GTF or GFF]] format to provide more context about the genes, such as the chromosome name, strand, and start and positions, and aid result interpretation.<ref name=sequence_analysis/><ref name=galaxy1/><ref name=galaxy2/><ref name=galaxy3>{{cite web |last1=Batut |first1=Bérénice |last2=Freeberg |first2=Mallory |last3=Heydarian |first3=Mo |display-authors=2 |title=Reference-based RNA-Seq data analysis (Galaxy Training Materials) |url=https://training.galaxyproject.org/training-material/topics/transcriptomics/tutorials/ref-based/tutorial.html |website=Galaxy Training! |date=17 March 2024 |access-date=26 April 2024}}</ref>

=== Functional enrichment analysis ===
{{See also|Transcriptomic technologies}}
Functional enrichment analysis identifies biological processes, pathways, and functional impacts associated with differentially expressed genes obtained from the previous step. It uses tools like GOSeq<ref>{{cite journal |last1=Young |first1=M.D. |last2=Wakefield |first2=M.J |last3=Smythe |first3=G.K |display-authors=2 |title=Gene ontology analysis for RNA-seq: accounting for selection bias |journal=Genome Biology |date=February 2010 |volume=11 |issue=2 |pages=R14 |doi=10.1186/gb-2010-11-2-r14 |doi-access=free |pmid=20132535 |hdl=11343/56416 |hdl-access=free |pmc=2872874 }}</ref> and Pathview.<ref>{{cite journal |last1=Luo |first1=Weijun |last2=Brouwer |first2=Cory |title=Pathview: an R/Bioconductor package for pathway-based data integration and visualization |journal=Bioinformatics |date=June 2013 |volume=29 |issue=14 |pages=1830–1831 |doi=10.1093/bioinformatics/btt285 |pmid=23740750 |pmc=3702256 }}</ref> This creates a table with information about what pathways and molecular processes are associated with the differentially expressed genes, what genes are down or upregulated, and what [[gene ontology]] terms are recurrent or over-represented.<ref name=sequence_analysis/><ref name=galaxy1/><ref name=galaxy2/><ref name=galaxy3/>

RNA sequence analysis explores gene expression dynamics and regulatory mechanisms underlying biological processes and diseases. Interpretation of images and tables are carried out within the context of the hypotheses being investigated.{{cn|date=August 2025}}

== Analyzing protein sequences ==

Proteome sequence analysis studies the complete set of proteins expressed by an organism or a cell under specific conditions. It describes protein structure, function, post-translational modifications, and interactions within biological systems. It often starts with raw [[mass spectrometry]] (MS) data from proteomics experiments, typically in mzML, mzXML, or RAW file formats.<ref name=sequence_analysis/>

Beyond preprocessing raw MS data to remove noise, normalize intensities, and detect peaks and converting proprietary file formats (e.g., RAW) to open-source formats (mzML, mzXML) for compatibility with downstream analysis tools, other analytical steps include [[peptide]] identification, peptide quantification, protein inference and quantification, generating quality control report, and normalization, imputation and significance testing. The choice and order of analytical steps depend on the MS method used, which can either be data dependent acquisition (DDA) or independent acquisition (DIA).<ref name=sequence_analysis/><ref>{{cite web |title=Quantitative mass spectrometry workflow |url=https://nf-co.re/quantms/1.2.0 |website=nf-core}}</ref>

== Genome browsers in sequence analysis ==

Genome browsers offer a non-code, user-friendly interface to visualize genomes and genomic segments, identify genomic features, and analyze the relationship between numerous genomic elements. The three primary genome browsers—Ensembl genome browser, UCSC genome browser, and the National Centre for Biotechnology Information (NCBI)—support different sequence analysis procedures, including genome assembly, genome annotation, and comparative genomics like exploring differential expression patterns and identifying conserved regions. All browsers support multiple data formats for upload and download and provide links to external tools and resources for sequence analyses, which contributes to their versatility.<ref>{{cite journal |last1=Wang |first1=Jun |last2=Kong |first2=Lei |last3=Gao |first3=Ge |display-authors=2 |title=A brief introduction to web-based genome browsers |journal=Briefings in Bioinformatics |date=July 2012 |volume=14 |issue=2 |pages=131–143 |doi=10.1093/bib/bbs029 |pmid=22764121 }}</ref><ref>{{cite book |last1=Pevsner |first1=J. |title=Bioinformatics for DNA Sequence Analysis |chapter=Analysis of Genomic DNA with the UCSC Genome Browser |series=Methods in Molecular Biology |date=February 2009 |volume=537 |pages=277–301 |publisher=Humana Press |doi=10.1007/978-1-59745-251-9_14 |pmid=19378150 |isbn=978-1-59745-251-9 |chapter-url=https://doi.org/10.1007/978-1-59745-251-9_14}}</ref>

== Sequence alignment ==
{{Main|Sequence_alignment#Pairwise_alignment}}
 {{Main|Multiple sequence alignment}}

[[File:WPP domain alignment.PNG|thumb|650px|center|Example multiple sequence alignment]]

There are millions of [[protein sequence|protein]] and [[nucleotide]] sequences known. These sequences fall into many groups of related sequences known as [[protein family|protein families]] or gene families.  Relationships between these sequences are usually discovered by aligning them together and assigning this alignment a score. There are two main types of sequence alignment. Pair-wise sequence alignment only compares two sequences at a time and multiple sequence alignment compares many sequences. Two important algorithms for aligning pairs of sequences are the [[Needleman-Wunsch algorithm]] and the [[Smith-Waterman algorithm]]. Popular tools for sequence alignment include:{{cn|date=August 2025}}
* Pair-wise alignment - [[BLAST (biotechnology)|BLAST]], [[Dot plot (bioinformatics)|Dot plots]]
* Multiple alignment - [[Clustal]]W, [[ProbCons|PROBCONS]], [[MUSCLE (alignment software)|MUSCLE]], [[MAFFT]], and  [[T-Coffee]].
A common use for pairwise sequence alignment is to take a sequence of interest and compare it to all known sequences in a database to identify [[Homology (biology)|homologous sequences]]. In general, the matches in the database are ordered to show the most closely related sequences first, followed by sequences with diminishing similarity. These matches are usually reported with a measure of statistical significance such as an [[Expectation value]].{{cn|date=August 2025}}

==Profile comparison==
In 1987, Michael Gribskov, Andrew McLachlan, and [[David Eisenberg]] introduced the method of profile comparison for identifying distant similarities between proteins.<ref name="pmid3474607">{{cite journal |author1=Gribskov M |author2=McLachlan AD |author3=Eisenberg D |title=Profile analysis: detection of distantly related proteins |journal=Proc. Natl. Acad. Sci. U.S.A. |volume=84 |issue=13 |pages=4355–8 |date=July 1987 |pmid=3474607 |pmc=305087 |doi= 10.1073/pnas.84.13.4355|bibcode = 1987PNAS...84.4355G |doi-access=free }}</ref> Rather than using a single sequence, profile methods use a multiple sequence alignment to encode a profile which contains information about the conservation level of each residue. These profiles can then be used to search collections of sequences to find sequences that are related. Profiles are also known as Position Specific Scoring Matrices (PSSMs). In 1993, a probabilistic interpretation of profiles was introduced by [[Anders Krogh]] and colleagues using [[hidden Markov models]].<ref name="pmid7584370">{{cite journal |author1=Brown M |author2=Hughey R |author3=Krogh A |author4=Mian IS |author5=Sjölander K |author6=Haussler D |title=Using Dirichlet mixture priors to derive hidden Markov models for protein families |journal=Proc Int Conf Intell Syst Mol Biol |volume=1 |pages=47–55 |year=1993 |pmid=7584370 }}</ref><ref name="pmid8107089">{{cite journal |author1=Krogh A |author2=Brown M |author3=Mian IS |author4=Sjölander K |author5=Haussler D |title=Hidden Markov models in computational biology. Applications to protein modeling |journal=J. Mol. Biol. |volume=235 |issue=5 |pages=1501–31 |date=February 1994 |pmid=8107089 |doi=10.1006/jmbi.1994.1104 |s2cid=2160404 }}</ref> These models have become known as profile-HMMs.

In recent years,{{When|date=February 2013}} methods have been developed that allow the comparison of profiles directly to each other. These are known as profile-profile comparison methods.<ref name="pmid21998158">{{cite journal |author1=Ye X |author2=Wang G |author3=Altschul SF |title=An assessment of substitution scores for protein profile-profile comparison |journal=Bioinformatics |volume=27 |issue=24 |pages=3356–63 |date=December 2011 |pmid=21998158 |doi=10.1093/bioinformatics/btr565 |pmc=3232366}}</ref>

== Sequence assembly ==

Sequence assembly refers to the reconstruction of a DNA sequence by [[sequence alignment|aligning]] and merging small DNA  fragments. It is an integral part of modern [[DNA sequencing]]. Since presently-available DNA sequencing technologies are ill-suited for reading long sequences, large pieces of DNA (such as genomes) are often sequenced by (1) cutting the DNA into small pieces, (2) reading the small fragments, and (3) reconstituting the original DNA by merging the information on various fragments.{{cn|date=August 2025}}

Recently, sequencing multiple species at one time is one of the top research objectives. Metagenomics is the study of microbial communities directly obtained from the environment. Different from cultured microorganisms from the lab, the wild sample usually contains dozens, sometimes even thousands of types of microorganisms from their original habitats.<ref name="A primer on metagenomics">{{cite journal|last1=Wooley|first1=JC|last2=Godzik|first2=A|last3=Friedberg|first3=I|title=A primer on metagenomics|journal=PLOS Comput Biol|date=Feb 26, 2010|volume=6|issue=2|doi=10.1371/journal.pcbi.1000667|pmid=20195499|pmc=2829047|article-number=e1000667|bibcode=2010PLSCB...6E0667W |doi-access=free }}</ref> Recovering the original genomes can prove to be very challenging.{{cn|date=August 2025}}

== Gene prediction ==
{{Main|Gene prediction}}
Gene prediction or gene finding refers to the process of identifying the regions of genomic DNA that encode [[genes]]. This includes protein-coding [[gene]]s as well as [[RNA gene]]s, but may also include the prediction of other functional elements such as [[regulatory regions]]. Geri is one of the first and most important steps in understanding the genome of a species once it has been [[Sequencing|sequenced]]. In general, the prediction of bacterial genes is significantly simpler and more accurate than the prediction of genes in eukaryotic species that usually have complex [[intron]]/[[exon]] patterns. Identifying genes in long sequences remains a problem, especially when the number of genes is unknown. [[Hidden markov models]] can be part of the solution.<ref>{{cite journal|last1=Stanke|first1=M|last2=Waack|first2=S|title=Gene prediction with a hidden Markov model and a new intron submodel|journal=Bioinformatics|volume=19 Suppl 2|date=Oct 19, 2003|issue=2|pages=215–25|pmid=14534192|doi=10.1093/bioinformatics/btg1080|doi-access=free}}</ref> Machine learning has played a significant role in predicting the sequence of transcription factors.<ref>{{cite journal|last1=Alipanahi|first1=B|last2=Delong|first2=A|last3=Weirauch|first3=MT|last4=Frey|first4=BJ|title=Predicting the sequence specificities of DNA- and RNA-binding proteins by deep learning|journal=Nat Biotechnol|date=Aug 2015|volume=33|issue=8|pages=831–8|doi=10.1038/nbt.3300|pmid=26213851|doi-access=free}}</ref> Traditional sequencing analysis focused on the statistical parameters of the nucleotide sequence itself (The most common programs used are listed in [https://www.ncbi.nlm.nih.gov/books/NBK20261/table/A181/?report=objectonly Table 4.1]). Another method is to identify homologous sequences based on other known gene sequences (Tools see [https://www.ncbi.nlm.nih.gov/books/NBK20261/table/A190/?report=objectonly Table 4.3]).<ref name="A primer on metagenomics"/>  The two methods described here are focused on the sequence. However, the shape feature of these molecules such as DNA and protein have also been studied and proposed to have an equivalent, if not higher, influence on the behaviors of these molecules.<ref>{{cite journal|last1=Abe|first1=N|last2=Dror|first2=I|last3=Yang|first3=L|last4=Slattery|first4=M|last5=Zhou|first5=T|last6=Bussemaker|first6=HJ|last7=Rohs R|first7=R|last8=Mann|first8=RS|title=Deconvolving the recognition of DNA shape from sequence|journal=Cell|date=Apr 9, 2015|volume=161|issue=2|pages=307–18|doi=10.1016/j.cell.2015.02.008|pmid=25843630|pmc=4422406}}</ref>

== Protein structure prediction ==
{{Main|Threading (protein sequence)}}

[[File:Target3dsmRib 354predictedModels CASP8.jpg|thumb|Target protein structure (3dsm, shown in ribbons), with Calpha backbones (in gray) of 354 predicted models for it submitted in the CASP8 structure-prediction experiment.|250px|right]]

The 3D structures of molecules are of major importance to their functions in nature. Since structural prediction of large molecules at an atomic level is a largely intractable problem, some biologists introduced ways to predict 3D structure at a primary sequence level. This includes the biochemical or statistical analysis of amino acid residues in local regions and structural the inference from homologs (or other potentially related proteins) with known 3D structures.{{cn|date=August 2025}}

There have been a large number of diverse approaches to solve the structure prediction problem.  In order to determine which methods were most effective, a structure prediction competition was founded called [[CASP]] (Critical Assessment of Structure Prediction).<ref name="pmid9485489">{{cite journal |author1=Moult J |author2=Hubbard T |author3=Bryant SH |author4=Fidelis K |author5=Pedersen JT |title=Critical assessment of methods of protein structure prediction (CASP): round II |journal=Proteins |volume=Suppl 1 |pages=2–6 |year=1997 |issue=S1 |pmid=9485489 |doi= 10.1002/(SICI)1097-0134(1997)1+<2::AID-PROT2>3.0.CO;2-T|s2cid=26823924 }}</ref>

== Computational approaches and techniques ==

Sequence analysis tasks are often non-trivial to resolve and require the use of relatively complex approaches, many of which are the backbone behind many existing sequence analysis tools. Of the many methods used in practice, the most popular include the following:{{cn|date=August 2025}}
* [[Dynamic programming]]
* [[Artificial neural network]]
* [[Hidden Markov model]]
* [[Support vector machine]]
* [[Cluster analysis|Clustering]]
* [[Bayesian network]]
* [[Regression analysis]]
* [[Sequence mining]]
* [[Alignment-free sequence analysis]]

==See also==
* [[List of sequence alignment software]]
* [[List of alignment visualization software]]
* [[List of phylogenetics software]]
* [[List of phylogenetic tree visualization software]]
* [[List of protein structure prediction software]]
* [[List of RNA structure prediction software]]
* [[Sequence analysis in social sciences#Software|Software section in Sequence analysis in social sciences]]

==References==
{{Reflist|2}}

{{DEFAULTSORT:Sequence Analysis}}
[[Category:Bioinformatics]]