CLUSTAL V is an improved version of the CLUSTAL package for multiple sequence alignment. It is a single program written in C, which can be used on any machine with a standard C compiler. The main new features include the ability to store and reuse old alignments and calculate phylogenetic trees after alignment. The program is simple to use, menu-driven, and provides on-line help. CLUSTAL V is portable to a wide variety of operating systems and has been rewritten in C. It retains the original CLUSTAL features of low core memory usage, speed, sensitivity, and capacity. The program supports flexible sequence input and output formats, and allows the calculation of phylogenetic trees with bootstrap confidence intervals after alignment. It can also align old alignments with each other. CLUSTAL V is compatible with various sequence formats, including NBRF/PIR, FASTA, and EMBL/Swissprot. It can be used on different platforms, including IBM PCs, Macintoshes, and Sun workstations. The program's performance depends on available memory, and it can handle hundreds of sequences on mainframes or workstations, while on smaller machines, speed becomes the limiting factor. The program is completely menu-driven, with on-line help available. It offers a command-line interface and allows users to change parameters and file names. It supports three different weight matrices for amino acid comparisons and can read a weight matrix from a file. Multiple alignments are performed in a progressive manner, using a guide tree constructed with the UPGMA method. The final alignment is achieved by aligning larger and larger alignments. The alignment algorithm used is a memory-efficient variation of the Gotoh algorithm. Phylogenetic trees are calculated using the Neighbor-Joining method, which does not assume a constant rate of evolution. Confidence intervals on the tree are calculated using a bootstrap procedure. Alignments can be written in various formats, including Phylip. Profile alignments allow the alignment of old alignments, which is useful for adding new sequences or predicting structures. The program is available from the EMBL file server and can be obtained via email or by sending a formatted diskette. It is widely used for multiple sequence alignment due to its portability, speed, and ease of use.CLUSTAL V is an improved version of the CLUSTAL package for multiple sequence alignment. It is a single program written in C, which can be used on any machine with a standard C compiler. The main new features include the ability to store and reuse old alignments and calculate phylogenetic trees after alignment. The program is simple to use, menu-driven, and provides on-line help. CLUSTAL V is portable to a wide variety of operating systems and has been rewritten in C. It retains the original CLUSTAL features of low core memory usage, speed, sensitivity, and capacity. The program supports flexible sequence input and output formats, and allows the calculation of phylogenetic trees with bootstrap confidence intervals after alignment. It can also align old alignments with each other. CLUSTAL V is compatible with various sequence formats, including NBRF/PIR, FASTA, and EMBL/Swissprot. It can be used on different platforms, including IBM PCs, Macintoshes, and Sun workstations. The program's performance depends on available memory, and it can handle hundreds of sequences on mainframes or workstations, while on smaller machines, speed becomes the limiting factor. The program is completely menu-driven, with on-line help available. It offers a command-line interface and allows users to change parameters and file names. It supports three different weight matrices for amino acid comparisons and can read a weight matrix from a file. Multiple alignments are performed in a progressive manner, using a guide tree constructed with the UPGMA method. The final alignment is achieved by aligning larger and larger alignments. The alignment algorithm used is a memory-efficient variation of the Gotoh algorithm. Phylogenetic trees are calculated using the Neighbor-Joining method, which does not assume a constant rate of evolution. Confidence intervals on the tree are calculated using a bootstrap procedure. Alignments can be written in various formats, including Phylip. Profile alignments allow the alignment of old alignments, which is useful for adding new sequences or predicting structures. The program is available from the EMBL file server and can be obtained via email or by sending a formatted diskette. It is widely used for multiple sequence alignment due to its portability, speed, and ease of use.