Genetic code

From Wikipedia, the free encyclopedia

A series of codons in part of a messenger RNA (mRNA) molecule. Each codon consists of three nucleotides, usually corresponding to a single amino acid. The nucleotides are abbreviated with the letters A, U, G and C. This is mRNA, which uses U (uracil). DNA uses T (thymine) instead. This mRNA molecule will instruct a ribosome to synthesize a protein according to this code.

The genetic code is the set of rules used by living cells to translate information encoded within genetic material (DNA or mRNA sequences) into proteins. Translation is accomplished by the ribosome, which links amino acids in an order specified by messenger RNA (mRNA), using transfer RNA (tRNA) molecules to carry amino acids and to read the mRNA three nucleotides at a time. The genetic code is highly similar among all organisms and can be expressed in a simple table with 64 entries.

The code defines how sequences of nucleotide triplets, called codons, specify which amino acid will be added next during protein synthesis. With some exceptions, a three-nucleotide codon in a nucleic acid sequence specifies a single amino acid. The vast majority of genes are encoded with a single scheme (see the RNA codon table). That scheme is often referred to as the canonical or standard genetic code, or simply the genetic code, though variant codes (such as in human mitochondria) exist.

While the "genetic code" determines a protein's amino acid sequence, other genomic regions determine when and where these proteins are produced according to various "gene regulatory codes".

History

The genetic code

 

Efforts to understand how proteins are encoded began after DNA's structure was discovered in 1953. George Gamow postulated that sets of three bases must be employed to encode the 20 standard amino acids used by living cells to build proteins, which would allow a maximum of 4³ = 64 amino acids.

Codons

The Crick, Brenner, Barnett and Watts-Tobin experiment first demonstrated that codons consist of three DNA bases. Marshall Nirenberg and Heinrich J. Matthaei were the first to reveal the nature of a codon in 1961.

They used a cell-free system to translate a poly-uracil RNA sequence (i.e., UUUUU...) and discovered that the polypeptide that they had synthesized consisted of only the amino acid phenylalanine. They thereby deduced that the codon UUU specified the amino acid phenylalanine.

This was followed by experiments in Severo Ochoa's laboratory that demonstrated that the poly-adenine RNA sequence (AAAAA...) coded for the polypeptide poly-lysine and that the poly-cytosine RNA sequence (CCCCC...) coded for the polypeptide poly-proline. Therefore, the codon AAA specified the amino acid lysine, and the codon CCC specified the amino acid proline. Using various copolymers most of the remaining codons were then determined.

Subsequent work by Har Gobind Khorana identified the rest of the genetic code. Shortly thereafter, Robert W. Holley determined the structure of transfer RNA (tRNA), the adapter molecule that facilitates the process of translating RNA into protein. This work was based upon Ochoa's earlier studies, yielding the latter the Nobel Prize in Physiology or Medicine in 1959 for work on the enzymology of RNA synthesis.

Extending this work, Nirenberg and Philip Leder revealed the code's triplet nature and deciphered its codons. In these experiments, various combinations of mRNA were passed through a filter that contained ribosomes, the components of cells that translate RNA into protein. Unique triplets promoted the binding of specific tRNAs to the ribosome. Leder and Nirenberg were able to determine the sequences of 54 out of 64 codons in their experiments. Khorana, Holley and Nirenberg received the 1968 Nobel for their work.

The three stop codons were named by discoverers Richard Epstein and Charles Steinberg. "Amber" was named after their friend Harris Bernstein, whose last name means "amber" in German. The other two stop codons were named "ochre" and "opal" in order to keep the "color names" theme.

Expanded genetic codes (synthetic biology)

In a broad academic audience, the concept of the evolution of the genetic code from the original and ambiguous genetic code to a well-defined ("frozen") code with the repertoire of 20 (+2) canonical amino acids is widely accepted. However, there are different opinions, concepts, approaches and ideas, which is the best way to change it experimentally. Even models are proposed that predict "entry points" for synthetic amino acid invasion of the genetic code.

Since 2001, 40 non-natural amino acids have been added into protein by creating a unique codon (recoding) and a corresponding transfer-RNA:aminoacyl – tRNA-synthetase pair to encode it with diverse physicochemical and biological properties in order to be used as a tool to exploring protein structure and function or to create novel or enhanced proteins.

H. Murakami and M. Sisido extended some codons to have four and five bases. Steven A. Benner constructed a functional 65th (in vivo) codon.

In 2015 N. Budisa, D. Söll and co-workers reported the full substitution of all 20,899 tryptophan residues (UGG codons) with unnatural thienopyrrole-alanine in the genetic code of the bacterium Escherichia coli.

In 2016 the first stable semisynthetic organism was created. It was a (single cell) bacterium with two synthetic bases (called X and Y). The bases survived cell division.

In 2017, researchers in South Korea reported that they had engineered a mouse with an extended genetic code that can produce proteins with unnatural amino acids.

Features

Reading frames in the DNA sequence of a region of the human mitochondrial genome coding for the genes MT-ATP8 and MT-ATP6 (in black: positions 8,525 to 8,580 in the sequence accession NC_012920). There are three possible reading frames in the 5' → 3' forward direction, starting on the first (+1), second (+2) and third position (+3). For each codon (square brackets), the amino acid is given by the vertebrate mitochondrial code, either in the +1 frame for MT-ATP8 (in red) or in the +3 frame for MT-ATP6 (in blue). The MT-ATP8 genes terminates with the TAG stop codon (red dot) in the +1 frame. The MT-ATP6 gene starts with the ATG codon (blue circle for the M amino acid) in the +3 frame.

Reading frame

A reading frame is defined by the initial triplet of nucleotides from which translation starts. It sets the frame for a run of successive, non-overlapping codons, which is known as an "open reading frame" (ORF). For example, the string 5'-AAATGAACG-3' (see figure), if read from the first position, contains the codons AAA, TGA, and ACG ; if read from the second position, it contains the codons AAT and GAA ; and if read from the third position, it contains the codons ATG and AAC. Every sequence can, thus, be read in its 5' → 3' direction in three reading frames, each producing a possibly distinct amino acid sequence: in the given example, Lys (K)-Trp (W)-Thr (T), Asn (N)-Glu (E), or Met (M)-Asn (N), respectively (when translating with the vertebrate mitochondrial code). When DNA is double-stranded, six possible reading frames are defined, three in the forward orientation on one strand and three reverse on the opposite strand. Protein-coding frames are defined by a start codon, usually the first AUG (ATG) codon in the RNA (DNA) sequence. 

In eukaryotes, ORFs in exons are often interrupted by introns.

Start/stop codons

Translation starts with a chain-initiation codon or start codon. The start codon alone is not sufficient to begin the process. Nearby sequences such as the Shine-Dalgarno sequence in E. coli and initiation factors are also required to start translation. The most common start codon is AUG, which is read as methionine or, in bacteria, as formylmethionine. Alternative start codons depending on the organism include "GUG" or "UUG"; these codons normally represent valine and leucine, respectively, but as start codons they are translated as methionine or formylmethionine.

The three stop codons have names: UAG is amber, UGA is opal (sometimes also called umber), and UAA is ochre. Stop codons are also called "termination" or "nonsense" codons. They signal release of the nascent polypeptide from the ribosome because no cognate tRNA has anticodons complementary to these stop signals, allowing a release factor to bind to the ribosome instead.

Effect of mutations

Examples of notable mutations that can occur in humans.

 

During the process of DNA replication, errors occasionally occur in the polymerization of the second strand. These errors, mutations, can affect an organism's phenotype, especially if they occur within the protein coding sequence of a gene. Error rates are typically 1 error in every 10–100 million bases—due to the "proofreading" ability of DNA polymerases.

Missense mutations and nonsense mutations are examples of point mutations that can cause genetic diseases such as sickle-cell disease and thalassemia respectively. Clinically important missense mutations generally change the properties of the coded amino acid residue among basic, acidic, polar or non-polar states, whereas nonsense mutations result in a stop codon.

Mutations that disrupt the reading frame sequence by indels (insertions or deletions) of a non-multiple of 3 nucleotide bases are known as frameshift mutations. These mutations usually result in a completely different translation from the original, and likely cause a stop codon to be read, which truncates the protein. These mutations may impair the protein's function and are thus rare in in vivo protein-coding sequences. One reason inheritance of frameshift mutations is rare is that, if the protein being translated is essential for growth under the selective pressures the organism faces, absence of a functional protein may cause death before the organism becomes viable. Frameshift mutations may result in severe genetic diseases such as Tay–Sachs disease.

Although most mutations that change protein sequences are harmful or neutral, some mutations have benefits. These mutations may enable the mutant organism to withstand particular environmental stresses better than wild type organisms, or reproduce more quickly. In these cases a mutation will tend to become more common in a population through natural selection. Viruses that use RNA as their genetic material have rapid mutation rates, which can be an advantage, since these viruses thereby evolve rapidly, and thus evade the immune system defensive responses. In large populations of asexually reproducing organisms, for example, E. coli, multiple beneficial mutations may co-occur. This phenomenon is called clonal interference and causes competition among the mutations.

Degeneracy

Degeneracy is the redundancy of the genetic code. This term was given by Bernfield and Nirenberg. The genetic code has redundancy but no ambiguity (see the codon tables below for the full correlation). For example, although codons GAA and GAG both specify glutamic acid (redundancy), neither specifies another amino acid (no ambiguity). The codons encoding one amino acid may differ in any of their three positions. For example, the amino acid leucine is specified by YUR or CUN (UUA, UUG, CUU, CUC, CUA, or CUG) codons (difference in the first or third position indicated using IUPAC notation), while the amino acid serine is specified by UCN or AGY (UCA, UCG, UCC, UCU, AGU, or AGC) codons (difference in the first, second, or third position). A practical consequence of redundancy is that errors in the third position of the triplet codon cause only a silent mutation or an error that would not affect the protein because the hydrophilicity or hydrophobicity is maintained by equivalent substitution of amino acids; for example, a codon of NUN (where N = any nucleotide) tends to code for hydrophobic amino acids. NCN yields amino acid residues that are small in size and moderate in hydropathy; NAN encodes average size hydrophilic residues. The genetic code is so well-structured for hydropathy that a mathematical analysis (Singular Value Decomposition) of 12 variables (4 nucleotides x 3 positions) yields a remarkable correlation (C = 0.95) for predicting the hydropathy of the encoded amino acid directly from the triplet nucleotide sequence, without translation. Note in the table, below, eight amino acids are not affected at all by mutations at the third position of the codon, whereas in the figure above, a mutation at the second position is likely to cause a radical change in the physicochemical properties of the encoded amino acid. Nevertheless, changes in the first position of the codons are more important than changes in the second position on a global scale. The reason may be that charge reversal (from a positive to a negative charge or vice versa) can only occur upon mutations in the first position, but never upon changes in the second position of a codon. Such charge reversal may have dramatic consequences for the structure or function of a protein. This aspect may have been largely underestimated by previous studies. 

Grouping of codons by amino acid residue molar volume and hydropathy. A more detailed version is available.

Codon usage bias

The frequency of codons, also known as codon usage bias, can vary from species to species with functional implications for the control of translation. The following codon usage table is for the human genome.

Standard codon tables

RNA codon table

Standard genetic code

1st base	2nd base								3rd base
	U		C		A		G
U	UUU	(Phe/F) Phenylalanine	UCU	(Ser/S) Serine	UAU	(Tyr/Y) Tyrosine	UGU	(Cys/C) Cysteine	U
	UUC		UCC		UAC		UGC		C
	UUA	(Leu/L) Leucine	UCA		UAA	Stop (Ochre)	UGA	Stop (Opal)	A
	UUG		UCG		UAG	Stop (Amber)	UGG	(Trp/W) Tryptophan	G
C	CUU		CCU	(Pro/P) Proline	CAU	(His/H) Histidine	CGU	(Arg/R) Arginine	U
	CUC		CCC		CAC		CGC		C
	CUA		CCA		CAA	(Gln/Q) Glutamine	CGA		A
	CUG		CCG		CAG		CGG		G
A	AUU	(Ile/I) Isoleucine	ACU	(Thr/T) Threonine	AAU	(Asn/N) Asparagine	AGU	(Ser/S) Serine	U
	AUC		ACC		AAC		AGC		C
	AUA		ACA		AAA	(Lys/K) Lysine	AGA	(Arg/R) Arginine	A
	AUG[A]	(Met/M) Methionine	ACG		AAG		AGG		G
G	GUU	(Val/V) Valine	GCU	(Ala/A) Alanine	GAU	(Asp/D) Aspartic acid	GGU	(Gly/G) Glycine	U
	GUC		GCC		GAC		GGC		C
	GUA		GCA		GAA	(Glu/E) Glutamic acid	GGA		A
	GUG		GCG		GAG		GGG		G

^A The codon AUG both codes for methionine and serves as an initiation site: the first AUG in an mRNA's coding region is where translation into protein begins.

^{B ^ ^ ^} The historical basis for designating the stop codons as amber, ochre and opal is described in an autobiography by Sydney Brenner and in a historical article by Bob Edgar.

Inverse table for the standard genetic code (compressed using IUPAC notation)

Amino acid	Codons	Compressed		Amino acid	Codons	Compressed
Ala / A	GCU, GCC, GCA, GCG	GCN		Leu / L	UUA, UUG, CUU, CUC, CUA, CUG	YUR, CUN
Arg / R	CGU, CGC, CGA, CGG, AGA, AGG	CGN, MGR		Lys / K	AAA, AAG	AAR
Asn / N	AAU, AAC	AAY		Met / M	AUG
Asp / D	GAU, GAC	GAY		Phe / F	UUU, UUC	UUY
Cys / C	UGU, UGC	UGY		Pro / P	CCU, CCC, CCA, CCG	CCN
Gln / Q	CAA, CAG	CAR		Ser / S	UCU, UCC, UCA, UCG, AGU, AGC	UCN, AGY
Glu / E	GAA, GAG	GAR		Thr / T	ACU, ACC, ACA, ACG	ACN
Gly / G	GGU, GGC, GGA, GGG	GGN		Trp / W	UGG
His / H	CAU, CAC	CAY		Tyr / Y	UAU, UAC	UAY
Ile / I	AUU, AUC, AUA	AUH		Val / V	GUU, GUC, GUA, GUG	GUN
START	AUG			STOP	UAA, UGA, UAG	UAR, URA

DNA codon table

The DNA codon table is essentially identical to that for RNA, but with U replaced by T.

Alternative genetic codes

Non-standard amino acids

In some proteins, non-standard amino acids are substituted for standard stop codons, depending on associated signal sequences in the messenger RNA. For example, UGA can code for selenocysteine and UAG can code for pyrrolysine. Selenocysteine became to be seen as the 21st amino acid, and pyrrolysine as the 22nd. Unlike selenocysteine, pyrrolysine-encoded UAG is translated with the participation of a dedicated aminoacyl-tRNA synthetase. Both selenocysteine and pyrrolysine may be present in the same organism. Although the genetic code is normally fixed in an organism, the achaeal prokaryote Acetohalobium arabaticum can expand its genetic code from 20 to 21 amino acids (by including pyrrolysine) under different conditions of growth.

Variations

Genetic code logo of the Globobulimina pseudospinescens mitochondrial genome. The logo shows the 64 codons from left to right, predicted alternatives in red (relative to the standard genetic code). Red line: stop codons. The height of each amino acid in the stack shows how often it is aligned to the codon in homologous protein domains. The stack height indicates the support for the prediction.

 

Variations on the standard code were predicted in the 1970s. The first was discovered in 1979, by researchers studying human mitochondrial genes. Many slight variants were discovered thereafter, including various alternative mitochondrial codes. These minor variants for example involve translation of the codon UGA as tryptophan in Mycoplasma species, and translation of CUG as a serine rather than leucine in yeasts of the "CTG clade" (such as Candida albicans). Because viruses must use the same genetic code as their hosts, modifications to the standard genetic code could interfere with viral protein synthesis or functioning. However, viruses such as totiviruses have adapted to the host's genetic code modification. In bacteria and archaea, GUG and UUG are common start codons. In rare cases, certain proteins may use alternative start codons. Surprisingly, variations in the interpretation of the genetic code exist also in human nuclear-encoded genes: In 2016, researchers studying the translation of malate dehydrogenase found that in about 4% of the mRNAs encoding this enzyme the stop codon is naturally used to encode the amino acids tryptophan and arginine. This type of recoding is induced by a high-readthrough stop codon context and it is referred to as functional translational readthrough.

Variant genetic codes used by an organism can be inferred by identifying highly conserved genes encoded in that genome, and comparing its codon usage to the amino acids in homologous proteins of other organisms. For example, the program FACIL infers a genetic code by searching which amino acids in homologous protein domains are most often aligned to every codon. The resulting amino acid probabilities for each codon are displayed in a genetic code logo, that also shows the support for a stop codon. 

Despite these differences, all known naturally occurring codes are very similar. The coding mechanism is the same for all organisms: three-base codons, tRNA, ribosomes, single direction reading and translating single codons into single amino acids.

Origin

The genetic code is a key part of the story of life, according to which self-replicating RNA molecules preceded life as we know it. The main hypothesis for life's origin is the RNA world hypothesis. Any model for the emergence of genetic code is intimately related to a model of the transfer from ribozymes (RNA enzymes) to proteins as the principal enzymes in cells. In line with the RNA world hypothesis, transfer RNA molecules appear to have evolved before modern aminoacyl-tRNA synthetases, so the latter cannot be part of the explanation of its patterns.

A hypothetical randomly evolved genetic code further motivates a biochemical or evolutionary model for its origin. If amino acids were randomly assigned to triplet codons, there would be 1.5 × 10⁸⁴ possible genetic codes. This number is found by calculating the number of ways that 21 items (20 amino acids plus one stop) can be placed in 64 bins, wherein each item is used at least once. However, the distribution of codon assignments in the genetic code is nonrandom. In particular, the genetic code clusters certain amino acid assignments. 

Amino acids that share the same biosynthetic pathway tend to have the same first base in their codons. This could be an evolutionary relic of an early, simpler genetic code with fewer amino acids that later evolved to code a larger set of amino acids. It could also reflect steric and chemical properties that had another effect on the codon during its evolution. Amino acids with similar physical properties also tend to have similar codons, reducing the problems caused by point mutations and mistranslations.

Given the non-random genetic triplet coding scheme, a tenable hypothesis for the origin of genetic code could address multiple aspects of the codon table, such as absence of codons for D-amino acids, secondary codon patterns for some amino acids, confinement of synonymous positions to third position, the small set of only 20 amino acids (instead of a number approaching 64), and the relation of stop codon patterns to amino acid coding patterns.

Three main hypotheses address the origin of the genetic code. Many models belong to one of them or to a hybrid:

• Random freeze: the genetic code was randomly created. For example, early tRNA-like ribozymes may have had different affinities for amino acids, with codons emerging from another part of the ribozyme that exhibited random variability. Once enough peptides were coded for, any major random change in the genetic code would have been lethal; hence it became "frozen".
• Stereochemical affinity: the genetic code is a result of a high affinity between each amino acid and its codon or anti-codon; the latter option implies that pre-tRNA molecules matched their corresponding amino acids by this affinity. Later during evolution, this matching was gradually replaced with matching by aminoacyl-tRNA synthetases.
• Optimality: the genetic code continued to evolve after its initial creation, so that the current code maximizes some fitness function, usually some kind of error minimization.

Hypotheses have addressed a variety of scenarios:

• Chemical principles govern specific RNA interaction with amino acids. Experiments with aptamers showed that some amino acids have a selective chemical affinity for their codons. Experiments showed that of 8 amino acids tested, 6 show some RNA triplet-amino acid association.
• Biosynthetic expansion. The genetic code grew from a simpler earlier code through a process of "biosynthetic expansion". Primordial life "discovered" new amino acids (for example, as by-products of metabolism) and later incorporated some of these into the machinery of genetic coding. Although much circumstantial evidence has been found to suggest that fewer amino acid types were used in the past, precise and detailed hypotheses about which amino acids entered the code in what order are controversial.
• Natural selection has led to codon assignments of the genetic code that minimize the effects of mutations. A recent hypothesis suggests that the triplet code was derived from codes that used longer than triplet codons (such as quadruplet codons). Longer than triplet decoding would increase codon redundancy and would be more error resistant. This feature could allow accurate decoding absent complex translational machinery such as the ribosome, such as before cells began making ribosomes.
• Information channels: Information-theoretic approaches model the process of translating the genetic code into corresponding amino acids as an error-prone information channel. The inherent noise (that is, the error) in the channel poses the organism with a fundamental question: how can a genetic code be constructed to withstand noise while accurately and efficiently translating information? These "rate-distortion" models suggest that the genetic code originated as a result of the interplay of the three conflicting evolutionary forces: the needs for diverse amino acids, for error-tolerance and for minimal resource cost. The code emerges at a transition when the mapping of codons to amino acids becomes nonrandom. The code's emergence is governed by the topology defined by the probable errors and is related to the map coloring problem.
• Game theory: Models based on signaling games combine elements of game theory, natural selection and information channels. Such models have been used to suggest that the first polypeptides were likely short and had non-enzymatic function. Game theoretic models suggested that the organization of RNA strings into cells may have been necessary to prevent "deceptive" use of the genetic code, i.e. preventing the ancient equivalent of viruses from overwhelming the RNA world.
• Stop codons: Codons for translational stops are also an interesting aspect to the problem of the origin of the genetic code. As an example for addressing stop codon evolution, it has been suggested that the stop codons are such that they are most likely to terminate translation early in the case of a frame shift error. In contrast, some stereochemical molecular models explain the origin of stop codons as "unassignable".

Compiler

From Wikipedia, the free encyclopedia

A compiler is a computer program that transforms computer code written in one programming language (the source language) into another programming language (the target language). Compilers are a type of translator that support digital devices, primarily computers. The name compiler is primarily used for programs that translate source code from a high-level programming language to a lower level language (e.g., assembly language, object code, or machine code) to create an executable program.

However, there are many different types of compilers. If the compiled program can run on a computer whose CPU or operating system is different from the one on which the compiler runs, the compiler is a cross-compiler. A bootstrap compiler is written in the language that it intends to compile. A program that translates from a low-level language to a higher level one is a decompiler. A program that translates between high-level languages is usually called a source-to-source compiler or transpiler. A language rewriter is usually a program that translates the form of expressions without a change of language. The term compiler-compiler refers to tools used to create parsers that perform syntax analysis.

A compiler is likely to perform many or all of the following operations: preprocessing, lexical analysis, parsing, semantic analysis (syntax-directed translation), conversion of input programs to an intermediate representation, code optimization and code generation. Compilers implement these operations in phases that promote efficient design and correct transformations of source input to target output. Program faults caused by incorrect compiler behavior can be very difficult to track down and work around; therefore, compiler implementers invest significant effort to ensure compiler correctness.

Compilers are not the only translators used to transform source programs. An interpreter is computer software that transforms and then executes the indicated operations. The translation process influences the design of computer languages which leads to a preference of compilation or interpretation. In practice, an interpreter can be implemented for compiled languages and compilers can be implemented for interpreted languages.

History

A diagram of the operation of a typical multi-language, multi-target compiler

 

Theoretical computing concepts developed by scientists, mathematicians, and engineers formed the basis of digital modern computing development during World War II. Primitive binary languages evolved because digital devices only understand ones and zeros and the circuit patterns in the underlying machine architecture. In the late 1940s, assembly languages were created to offer a more workable abstraction of the computer architectures. Limited memory capacity of early computers led to substantial technical challenges when the first compilers were designed. Therefore, the compilation process needed to be divided into several small programs. The front end programs produce the analysis products used by the back end programs to generate target code. As computer technology provided more resources, compiler designs could align better with the compilation process. 

It is usually more productive for a programmer to use a high-level language, so the development of high-level languages followed naturally from the capabilities offered by digital computers. High-level languages are formal languages that are strictly defined by their syntax and semantics which form the high-level language architecture. Elements of these formal languages include:

• Alphabet, any finite set of symbols;
• String, a finite sequence of symbols;
• Language, any set of strings on an alphabet.

The sentences in a language may be defined by a set of rules called a grammar.

Backus–Naur form (BNF) describes the syntax of "sentences" of a language and was used for the syntax of Algol 60 by John Backus. The ideas derive from the context-free grammar concepts by Noam Chomsky, a linguist. "BNF and its extensions have become standard tools for describing the syntax of programming notations, and in many cases parts of compilers are generated automatically from a BNF description."

In the 1940s, Konrad Zuse designed an algorithmic programming language called Plankalkül ("Plan Calculus"). While no actual implementation occurred until the 1970s, it presented concepts later seen in APL designed by Ken Iverson in the late 1950s. APL is a language for mathematical computations. 

High-level language design during the formative years of digital computing provided useful programming tools for a variety of applications:

• FORTRAN (Formula Translation) for engineering and science applications is considered to be the first high-level language.
• COBOL (Common Business-Oriented Language) evolved from A-0 and FLOW-MATIC to become the dominant high-level language for business applications.
• LISP (List Processor) for symbolic computation.

Compiler technology evolved from the need for a strictly defined transformation of the high-level source program into a low-level target program for the digital computer. The compiler could be viewed as a front end to deal with the analysis of the source code and a back end to synthesize the analysis into the target code. Optimization between the front end and back end could produce more efficient target code.

Some early milestones in the development of compiler technology:

• 1952 – An Autocode compiler developed by Alick Glennie for the Manchester Mark I computer at the University of Manchester is considered by some to be the first compiled programming language.
• 1952 – Grace Hopper's team at Remington Rand wrote the compiler for the A-0 programming language (and coined the term compiler to describe it), although the A-0 compiler functioned more as a loader or linker than the modern notion of a full compiler.
• 1954-1957 – A team led by John Backus at IBM developed FORTRAN which is usually considered the first high-level language. In 1957, they completed a FORTRAN compiler that is generally credited as having introduced the first unambiguously complete compiler.
• 1959 – The Conference on Data Systems Language (CODASYL) initiated development of COBOL. The COBOL design drew on A-0 and FLOW-MATIC. By the early 1960s COBOL was compiled on multiple architectures.
• 1958-1962 – John McCarthy at MIT designed LISP. The symbol processing capabilities provided useful features for artificial intelligence research. In 1962, LISP 1.5 release noted some tools: an interpreter written by Stephen Russell and Daniel J. Edwards, a compiler and assembler written by Tim Hart and Mike Levin.

Early operating systems and software were written in assembly language. In the 60s and early 70s, the use of high-level languages for system programming was still controversial due to resource limitations. However, several research and industry efforts began the shift toward high-level systems programming languages, for example, BCPL, BLISS, B, and C. 

BCPL (Basic Combined Programming Language) designed in 1966 by Martin Richards at the University of Cambridge was originally developed as a compiler writing tool. Several compilers have been implemented, Richards' book provides insights to the language and its compiler. BCPL was not only an influential systems programming language that is still used in research but also provided a basis for the design of B and C languages. 

BLISS (Basic Language for Implementation of System Software) was developed for a Digital Equipment Corporation (DEC) PDP-10 computer by W.A. Wulf's Carnegie Mellon University (CMU) research team. The CMU team went on to develop BLISS-11 compiler one year later in 1970.

Multics (Multiplexed Information and Computing Service), a time-sharing operating system project, involved MIT, Bell Labs, General Electric (later Honeywell) and was led by Fernando Corbató from MIT. Multics was written in the PL/I language developed by IBM and IBM User Group. IBM's goal was to satisfy business, scientific, and systems programming requirements. There were other languages that could have been considered but PL/I offered the most complete solution even though it had not been implemented. For the first few years of the Mulitics project, a subset of the language could be compiled to assembly language with the Early PL/I (EPL) compiler by Doug McIlory and Bob Morris from Bell Labs. EPL supported the project until a boot-strapping compiler for the full PL/I could be developed.

Bell Labs left the Multics project in 1969: "Over time, hope was replaced by frustration as the group effort initially failed to produce an economically useful system." Continued participation would drive up project support costs. So researchers turned to other development efforts. A system programming language B based on BCPL concepts was written by Dennis Ritchie and Ken Thompson. Ritchie created a boot-strapping compiler for B and wrote Unics (Uniplexed Information and Computing Service) operating system for a PDP-7 in B. Unics eventually became spelled Unix.

Bell Labs started development and expansion of C based on B and BCPL. The BCPL compiler had been transported to Multics by Bell Labs and BCPL was a preferred language at Bell Labs. Initially, a front-end program to Bell Labs' B compiler was used while a C compiler was developed. In 1971, a new PDP-11 provided the resource to define extensions to B and rewrite the compiler. By 1973 the design of C language was essentially complete and the Unix kernel for a PDP-11 was rewritten in C. Steve Johnson started development of Portable C Compiler (PCC) to support retargeting of C compilers to new machines.

Object-oriented programming (OOP) offered some interesting possibilities for application development and maintenance. OOP concepts go further back but were part of LISP and Simula language science. At Bell Labs, the development of C++ became interested in OOP. C++ was first used in 1980 for systems programming. The initial design leveraged C language systems programming capabilities with Simula concepts. Object-oriented facilities were added in 1983. The Cfront program implemented a C++ front-end for C84 language compiler. In subsequent years several C++ compilers were developed as C++ popularity grew. 

In many application domains, the idea of using a higher-level language quickly caught on. Because of the expanding functionality supported by newer programming languages and the increasing complexity of computer architectures, compilers became more complex.

DARPA (Defense Advanced Research Projects Agency) sponsored a compiler project with Wulf's CMU research team in 1970. The Production Quality Compiler-Compiler PQCC design would produce a Production Quality Compiler (PQC) from formal definitions of source language and the target. PQCC tried to extend the term compiler-compiler beyond the traditional meaning as a parser generator (e.g., Yacc) without much success. PQCC might more properly be referred to as a compiler generator.

PQCC research into code generation process sought to build a truly automatic compiler-writing system. The effort discovered and designed the phase structure of the PQC. The BLISS-11 compiler provided the initial structure. The phases included analyses (front end), intermediate translation to virtual machine (middle end), and translation to the target (back end). TCOL was developed for the PQCC research to handle language specific constructs in the intermediate representation. Variations of TCOL supported various languages. The PQCC project investigated techniques of automated compiler construction. The design concepts proved useful in optimizing compilers and compilers for the object-oriented programming language Ada. 

The Ada Stoneman Document formalized the program support environment (APSE) along with the kernel (KAPSE) and minimal (MAPSE). An Ada interpreter NYU/ED supported development and standardization efforts with the American National Standards Institute (ANSI) and the International Standards Organization (ISO). Initial Ada compiler development by the U.S. Military Services included the compilers in a complete integrated design environment along the lines of the Stoneman Document. Army and Navy worked on the Ada Language System (ALS) project targeted to DEC/VAX architecture while the Air Force started on the Ada Integrated Environment (AIE) targeted to IBM 370 series. While the projects did not provide the desired results, they did contribute to the overal effort on Ada development.

Other Ada compiler efforts got underway in Britain at the University of York and in Germany at the University of Karlsruhe. In the U. S., Verdix (later acquired by Rational) delivered the Verdix Ada Development System (VADS) to the Army. VADS provided a set of development tools including a compiler. Unix/VADS could be hosted on a variety of Unix platforms such as DEC Ultrix and the Sun 3/60 Solaris targeted to Motorola 68020 in an Army CECOM evaluation. There were soon many Ada compilers available that passed the Ada Validation tests. The Free Software Foundation GNU project developed the GNU Compiler Collection (GCC) which provides a core capability to support multiple languages and targets. The Ada version GNAT is one of the most widely used Ada compilers. GNAT is free but there is also commercial support, for example, AdaCore, was founded in 1994 to provide commercial software solutions for Ada. GNAT Pro includes the GNU GCC based GNAT with a tool suite to provide an integrated development environment. 

High-level languages continued to drive compiler research and development. Focus areas included optimization and automatic code generation. Trends in programming languages and development environments influenced compiler technology. More compilers became included in language distributions (PERL, Java Development Kit) and as a component of an IDE (VADS, Eclipse, Ada Pro). The interrelationship and interdependence of technologies grew. The advent of web services promoted growth of web languages and scripting languages. Scripts trace back to the early days of Command Line Interfaces (CLI) where the user could enter commands to be executed by the system. User Shell concepts developed with languages to write shell programs. Early Windows designs offered a simple batch programming capability. The conventional transformation of these language used an interpreter. While not widely used, Bash and Batch compilers have been written. More recently sophisticated interpreted languages became part of the developers tool kit. Modern scripting languages include PHP, Python, Ruby and Lua. (Lua is widely used in game development.) All of these have interpreter and compiler support.

"When the field of compiling began in the late 50s, its focus was limited to the translation of high-level language programs into machine code ... The compiler field is increasingly intertwined with other disciplines including computer architecture, programming languages, formal methods, software engineering, and computer security." The "Compiler Research: The Next 50 Years" article noted the importance of object-oriented languages and Java. Security and parallel computing were cited among the future research targets.

Compiler construction

A compiler implements a formal transformation from a high-level source program to a low-level target program. Compiler design can define an end to end solution or tackle a defined subset that interfaces with other compilation tools e.g. preprocessors, assemblers, linkers. Design requirements include rigorously defined interfaces both internally between compiler components and externally between supporting toolsets. 

In the early days, the approach taken to compiler design was directly affected by the complexity of the computer language to be processed, the experience of the person(s) designing it, and the resources available. Resource limitations led to the need to pass through the source code more than once. 

A compiler for a relatively simple language written by one person might be a single, monolithic piece of software. However, as the source language grows in complexity the design may be split into a number of interdependent phases. Separate phases provide design improvements that focus development on the functions in the compilation process.

One-pass versus multi-pass compilers

Classifying compilers by number of passes has its background in the hardware resource limitations of computers. Compiling involves performing lots of work and early computers did not have enough memory to contain one program that did all of this work. So compilers were split up into smaller programs which each made a pass over the source (or some representation of it) performing some of the required analysis and translations. 

The ability to compile in a single pass has classically been seen as a benefit because it simplifies the job of writing a compiler and one-pass compilers generally perform compilations faster than multi-pass compilers. Thus, partly driven by the resource limitations of early systems, many early languages were specifically designed so that they could be compiled in a single pass (e.g., Pascal).

In some cases the design of a language feature may require a compiler to perform more than one pass over the source. For instance, consider a declaration appearing on line 20 of the source which affects the translation of a statement appearing on line 10. In this case, the first pass needs to gather information about declarations appearing after statements that they affect, with the actual translation happening during a subsequent pass. 

The disadvantage of compiling in a single pass is that it is not possible to perform many of the sophisticated optimizations needed to generate high quality code. It can be difficult to count exactly how many passes an optimizing compiler makes. For instance, different phases of optimization may analyse one expression many times but only analyse another expression once.

Splitting a compiler up into small programs is a technique used by researchers interested in producing provably correct compilers. Proving the correctness of a set of small programs often requires less effort than proving the correctness of a larger, single, equivalent program.

Three-stage compiler structure

Compiler design

Regardless of the exact number of phases in the compiler design, the phases can be assigned to one of three stages. The stages include a front end, a middle end, and a back end.

• The front end verifies syntax and semantics according to a specific source language. For statically typed languages it performs type checking by collecting type information. If the input program is syntactically incorrect or has a type error, it generates errors and warnings, highlighting them on the source code. Aspects of the front end include lexical analysis, syntax analysis, and semantic analysis. The front end transforms the input program into an intermediate representation (IR) for further processing by the middle end. This IR is usually a lower-level representation of the program with respect to the source code.
• The middle end performs optimizations on the IR that are independent of the CPU architecture being targeted. This source code/machine code independence is intended to enable generic optimizations to be shared between versions of the compiler supporting different languages and target processors. Examples of middle end optimizations are removal of useless (dead code elimination) or unreachable code (reachability analysis), discovery and propagation of constant values (constant propagation), relocation of computation to a less frequently executed place (e.g., out of a loop), or specialization of computation based on the context. Eventually producing the "optimized" IR that is used by the back end.
• The back end takes the optimized IR from the middle end. It may perform more analysis, transformations and optimizations that are specific for the target CPU architecture. The back end generates the target-dependent assembly code, performing register allocation in the process. The back end performs instruction scheduling, which re-orders instructions to keep parallel execution units busy by filling delay slots. Although most algorithms for optimization are NP-hard, heuristic techniques are well-developed and currently implemented in production-quality compilers. Typically the output of a back end is machine code specialized for a particular processor and operating system.

This front/middle/back-end approach makes it possible to combine front ends for different languages with back ends for different CPUs while sharing the optimizations of the middle end. Practical examples of this approach are the GNU Compiler Collection, LLVM, and the Amsterdam Compiler Kit, which have multiple front-ends, shared optimizations and multiple back-ends.

Front end

Lexer and parser example for C. Starting from the sequence of characters "if(net>0.0)total+=net*(1.0+tax/100.0);", the scanner composes a sequence of tokens, and categorizes each of them, for example as identifier, reserved word, number literal, or operator. The latter sequence is transformed by the parser into a syntax tree, which is then treated by the remaining compiler phases. The scanner and parser handles the regular and properly context-free parts of the grammar for C, respectively.

 

The front end analyzes the source code to build an internal representation of the program, called the intermediate representation (IR). It also manages the symbol table, a data structure mapping each symbol in the source code to associated information such as location, type and scope.

While the frontend can be a single monolithic function or program, as in a scannerless parser, it is more commonly implemented and analyzed as several phases, which may execute sequentially or concurrently. This method is favored due to its modularity and separation of concerns. Most commonly today, the frontend is broken into three phases: lexical analysis (also known as lexing), syntax analysis (also known as scanning or parsing), and semantic analysis. Lexing and parsing comprise the syntactic analysis (word syntax and phrase syntax, respectively), and in simple cases these modules (the lexer and parser) can be automatically generated from a grammar for the language, though in more complex cases these require manual modification. The lexical grammar and phrase grammar are usually context-free grammars, which simplifies analysis significantly, with context-sensitivity handled at the semantic analysis phase. The semantic analysis phase is generally more complex and written by hand, but can be partially or fully automated using attribute grammars. These phases themselves can be further broken down: lexing as scanning and evaluating, and parsing as building a concrete syntax tree (CST, parse tree) and then transforming it into an abstract syntax tree (AST, syntax tree). In some cases additional phases are used, notably line reconstruction and preprocessing, but these are rare. 

The main phases of the front end include the following:

• Line reconstruction converts the input character sequence to a canonical form ready for the parser. Languages which strop their keywords or allow arbitrary spaces within identifiers require this phase. The top-down, recursive-descent, table-driven parsers used in the 1960s typically read the source one character at a time and did not require a separate tokenizing phase. Atlas Autocode and Imp (and some implementations of ALGOL and Coral 66) are examples of stropped languages whose compilers would have a Line Reconstruction phase.
• Preprocessing supports macro substitution and conditional compilation. Typically the preprocessing phase occurs before syntactic or semantic analysis; e.g. in the case of C, the preprocessor manipulates lexical tokens rather than syntactic forms. However, some languages such as Scheme support macro substitutions based on syntactic forms.
• Lexical analysis (also known as lexing or tokenization) breaks the source code text into a sequence of small pieces called lexical tokens. This phase can be divided into two stages: the scanning, which segments the input text into syntactic units called lexemes and assign them a category; and the evaluating, which converts lexemes into a processed value. A token is a pair consisting of a token name and an optional token value. Common token categories may include identifiers, keywords, separators, operators, literals and comments, although the set of token categories varies in different programming languages. The lexeme syntax is typically a regular language, so a finite state automaton constructed from a regular expression can be used to recognize it. The software doing lexical analysis is called a lexical analyzer. This may not be a separate step—it can be combined with the parsing step in scannerless parsing, in which case parsing is done at the character level, not the token level.
• Syntax analysis (also known as parsing) involves parsing the token sequence to identify the syntactic structure of the program. This phase typically builds a parse tree, which replaces the linear sequence of tokens with a tree structure built according to the rules of a formal grammar which define the language's syntax. The parse tree is often analyzed, augmented, and transformed by later phases in the compiler.
• Semantic analysis adds semantic information to the parse tree and builds the symbol table. This phase performs semantic checks such as type checking (checking for type errors), or object binding (associating variable and function references with their definitions), or definite assignment (requiring all local variables to be initialized before use), rejecting incorrect programs or issuing warnings. Semantic analysis usually requires a complete parse tree, meaning that this phase logically follows the parsing phase, and logically precedes the code generation phase, though it is often possible to fold multiple phases into one pass over the code in a compiler implementation.

Middle end

The middle end performs optimizations on the intermediate representation in order to improve the performance and the quality of the produced machine code. The middle end contains those optimizations that are independent of the CPU architecture being targeted. 

The main phases of the middle end include the following:

• Analysis: This is the gathering of program information from the intermediate representation derived from the input; data-flow analysis is used to build use-define chains, together with dependence analysis, alias analysis, pointer analysis, escape analysis, etc. Accurate analysis is the basis for any compiler optimization. The control flow graph of every compiled function and the call graph of the program are usually also built during the analysis phase.
• Optimization: the intermediate language representation is transformed into functionally equivalent but faster (or smaller) forms. Popular optimizations are inline expansion, dead code elimination, constant propagation, loop transformation and even automatic parallelization.

Compiler analysis is the prerequisite for any compiler optimization, and they tightly work together. For example, dependence analysis is crucial for loop transformation.

The scope of compiler analysis and optimizations vary greatly; their scope may range from operating within a basic block, to whole procedures, or even the whole program. There is a trade-off between the granularity of the optimizations and the cost of compilation. For example, peephole optimizations are fast to perform during compilation but only affect a small local fragment of the code, and can be performed independently of the context in which the code fragment appears. In contrast, interprocedural optimization requires more compilation time and memory space, but enable optimizations which are only possible by considering the behavior of multiple functions simultaneously.

Interprocedural analysis and optimizations are common in modern commercial compilers from HP, IBM, SGI, Intel, Microsoft, and Sun Microsystems. The free software GCC was criticized for a long time for lacking powerful interprocedural optimizations, but it is changing in this respect. Another open source compiler with full analysis and optimization infrastructure is Open64, which is used by many organizations for research and commercial purposes.

Due to the extra time and space needed for compiler analysis and optimizations, some compilers skip them by default. Users have to use compilation options to explicitly tell the compiler which optimizations should be enabled.

Back end

The back end is responsible for the CPU architecture specific optimizations and for code generation.

The main phases of the back end include the following:

• Machine dependent optimizations: optimizations that depend on the details of the CPU architecture that the compiler targets. A prominent example is peephole optimizations, which rewrites short sequences of assembler instructions into more efficient instructions.
• Code generation: the transformed intermediate language is translated into the output language, usually the native machine language of the system. This involves resource and storage decisions, such as deciding which variables to fit into registers and memory and the selection and scheduling of appropriate machine instructions along with their associated addressing modes (see also Sethi-Ullman algorithm). Debug data may also need to be generated to facilitate debugging.

Compiler correctness

Compiler correctness is the branch of software engineering that deals with trying to show that a compiler behaves according to its language specification. Techniques include developing the compiler using formal methods and using rigorous testing (often called compiler validation) on an existing compiler.

Compiled versus interpreted languages

Higher-level programming languages usually appear with a type of translation in mind: either designed as compiled language or interpreted language. However, in practice there is rarely anything about a language that requires it to be exclusively compiled or exclusively interpreted, although it is possible to design languages that rely on re-interpretation at run time. The categorization usually reflects the most popular or widespread implementations of a language — for instance, BASIC is sometimes called an interpreted language, and C a compiled one, despite the existence of BASIC compilers and C interpreters. 

Interpretation does not replace compilation completely. It only hides it from the user and makes it gradual. Even though an interpreter can itself be interpreted, a directly executed program is needed somewhere at the bottom of the stack. 

Further, compilers can contain interpreters for optimization reasons. For example, where an expression can be executed during compilation and the results inserted into the output program, then it prevents it having to be recalculated each time the program runs, which can greatly speed up the final program. Modern trends toward just-in-time compilation and bytecode interpretation at times blur the traditional categorizations of compilers and interpreters even further.

Some language specifications spell out that implementations must include a compilation facility; for example, Common Lisp. However, there is nothing inherent in the definition of Common Lisp that stops it from being interpreted. Other languages have features that are very easy to implement in an interpreter, but make writing a compiler much harder; for example, APL, SNOBOL4, and many scripting languages allow programs to construct arbitrary source code at runtime with regular string operations, and then execute that code by passing it to a special evaluation function. To implement these features in a compiled language, programs must usually be shipped with a runtime library that includes a version of the compiler itself.

Types

One classification of compilers is by the platform on which their generated code executes. This is known as the target platform.

 

A native or hosted compiler is one whose output is intended to directly run on the same type of computer and operating system that the compiler itself runs on. The output of a cross compiler is designed to run on a different platform. Cross compilers are often used when developing software for embedded systems that are not intended to support a software development environment. 

The output of a compiler that produces code for a virtual machine (VM) may or may not be executed on the same platform as the compiler that produced it. For this reason such compilers are not usually classified as native or cross compilers.

The lower level language that is the target of a compiler may itself be a high-level programming language. C, often viewed as some sort of portable assembler, can also be the target language of a compiler. E.g.: Cfront, the original compiler for C++ used C as target language. The C created by such a compiler is usually not intended to be read and maintained by humans. So indent style and pretty C intermediate code are irrelevant. Some features of C turn it into a good target language. E.g.: C code with #line directives can be generated to support debugging of the original source. 

While a common compiler type outputs machine code, there are many other types:

• A source-to-source compiler is a type of compiler that takes a high-level language as its input and outputs a high-level language. For example, an automatic parallelizing compiler will frequently take in a high-level language program as an input and then transform the code and annotate it with parallel code annotations (e.g. OpenMP) or language constructs (e.g. Fortran's DOALL statements).
• Bytecode compilers that compile to assembly language of a theoretical machine, like some Prolog implementations
  • This Prolog machine is also known as the Warren Abstract Machine (or WAM).
  • Bytecode compilers for Java, Python are also examples of this category.
• A Just-in-time compiler (JIT compiler) defers compilation until runtime. JIT compilers exist for many modern languages including Python, Javascript, Smalltalk, Java, Microsoft .NET's Common Intermediate Language (CIL) and others. A JIT compiler generally runs inside an interpreter. When the interpreter detects that a code path is "hot", meaning it is executed frequently, the JIT compiler will be invoked and compile the "hot" code for increased performance.
  • For some languages, such as Java, applications are first compiled using a bytecode compiler and delivered in a machine-independent intermediate representation. A bytecode interpreter executes the bytecode, but the JIT compiler will translate the bytecode to machine code when increased performance is necessary.
• hardware compilers (also known as syntheses tools) are compilers whose output is a description of the hardware configuration instead of a sequence of instructions.
  • The output of these compilers target computer hardware at a very low level, for example a field-programmable gate array (FPGA) or structured application-specific integrated circuit (ASIC). Such compilers are said to be hardware compilers, because the source code they compile effectively controls the final configuration of the hardware and how it operates. The output of the compilation is only an interconnection of transistors or lookup tables.
  • An example of hardware compiler is XST, the Xilinx Synthesis Tool used for configuring FPGAs. Similar tools are available from Altera, Synplicity, Synopsys and other hardware vendors.
• An assembler is a program that compiles human readable assembly language to machine code, the actual instructions executed by hardware. The inverse program that translates machine code to assembly language is called a disassembler.
• A program that translates from a low-level language to a higher level one is a decompiler.
• A program that translates between high-level languages is usually called a language translator, source-to-source compiler, language converter, or language rewriter.^{[citation needed]} The last term is usually applied to translations that do not involve a change of language.
• A program that translates into an object code format that is not supported on the compilation machine is called a cross compiler and is commonly used to prepare code for embedded applications.
• A program that rewrites object code back into the same type of object code while applying optimisations and transformations is a binary recompiler.