Skip to main content

​
Overview

A3M is the primary alignment format used by HH-suite tools. It is similar to A2M format but with a more compact representation of insertions and deletions.

​
Format Specification

​
Character Encoding

  • Upper case letters (A-Z): Match states (aligned columns)
  • Lower case letters (a-z): Insert states (insertions relative to match states)
  • Dash (-): Deletion in the match state
  • Dot (.): Gap aligned to an insert state (optional, may be omitted)

​
File Structure

An A3M file consists of:
  1. Header lines: Begin with > followed by sequence identifier and description
  2. Sequence lines: Aligned sequences using the character encoding above

​
Example

>sp|Q5VUD6|FA69B_HUMAN Protein FAM69B OS=Homo sapiens GN=FAM69B PE=2 SV=3
MRRLRRLAHLVLFCPFSKRLQGRLPGLRVRCIFLAWLGVFAGSWLVYVHYSSYSERCRGHVCQVVICDQYRKGIISGSVCQDLCELHMVEWRTCLSVAPG
QQVYSGLWRDKDVTIKCGIEETLDSKARSDAAPRRELVLFDKPTRGTSIKEFREMTLSFLKANLGDLPSLPALVGQVLLMADFNKDNRVSLAEAKSVWAL
LQRNEFLLLLSLQEKEHASRLLGYCGDLYLTEGVPHGAWHAAALPPLLRPLLPPALQGALQQWLGPAWPWRAKIAIGLLEFVEELFHGSYGTFYMCETTL
ANVGYTATYDFKMADLQQVAPEATVRRFLQGRRCEHSTDCTYGRDCRAPCDRLMRQCKGDLIQPNLAKVCALLRGYLLPGAPADLREELGTQLRTCTTLS
GLASQVEAHHSLVLSHLKTLLWKKISNTKYS
>tr|Q4S137|Q4S137_TETNG Chromosome 1 SCAF14770, whole genome shotgun sequence OS=Tetraodon nigroviridis GN=GSTENG00025741001 PE=4 SV=1
------------------YvqrkesgiegplgsratdgraggLDARFSYLHMKYLFLSWLAVFVGSWVVYVEYSSYTELCRGRECKNSIvssfiflirlfqrrglpvahllsvlskCDKYRRGLIDGSACSSLCEKGTLSLGTCFSARAKSQVYTGSWGDLEGVIKCRMEEAQRYDLENQ-------------------------------SKVGDQANLADLAAQVLSMTDANKDGHISLPEAHSTWALLQLNEFLLALVLQDREHTPKLLGFCGDLYVTEKVPYSPLYGVGLPWILEVWIPASLRHSMDQWFTPSWPFKAKISIGLLELVEDVFHGTFGSFLMCDVSAGSFGYNDRHDLKVTDARYIVPEAVFQEDIRQQRCDDDRDCRFGADCLTSCDLTKHRCTTEVTRPNLAKACETMKNYVLRGAPPDVREELEKQLYACMALKGSAEQMEMEHSLILNNLKTLLWKRISHTKDS

​
Key Features

​
Match vs Insert States

The distinction between match states (upper case) and insert states (lower case) is crucial:
  • Match states: Form the consensus structure of the alignment
  • Insert states: Represent insertions that are not part of the core alignment

​
Gaps

  • Deletions in match states are represented by -
  • Gaps aligned to insert states can be represented by . but are often omitted for compactness

​
Conversion from FASTA

When converting from FASTA format, you can specify how match states are assigned: 
# Using -M a3m: Upper/lower case in input determines match/insert
hhblits -i input.a3m -d database

# Using -M first: First sequence determines match columns
hhblits -i input.fasta -M first -d database

# Using -M [0-100]: Columns with <X% gaps are match states
hhblits -i input.fasta -M 50 -d database

​
Tools that Use A3M

A3M format is used by:
  • hhblits: For input queries and output MSAs
  • hhsearch: For input queries
  • hhalign: For input alignments
  • hhmake: For building HMMs
  • hhfilter: For filtering alignments
  • hhconsensus: For consensus calculation

​
Best Practices

​
Header Lines

  • Include informative sequence identifiers
  • Add organism and gene information when available
  • Use standard format: >identifier description

​
Sequence Quality

  • Ensure proper upper/lower case for match/insert states
  • Remove sequences with too many gaps using hhfilter
  • Filter redundant sequences to improve diversity

​
File Size

A3M files can be large. To reduce size:
  • Filter sequences by identity: hhfilter -id 90
  • Remove low-coverage sequences: hhfilter -cov 50
  • Select diverse sequences: hhfilter -diff 1000

​
Comparison with A2M

Key differences:
FeatureA2MA3M
Insert gapsMust use .Can be omitted
CompactnessLess compactMore compact
UsageLegacy formatCurrent standard

​
See Also

Build docs developers (and LLMs) love

Get started for freeTalk to us