Command-Line Parameters

Overview

HH-suite tools share many common parameters while also providing tool-specific options. This guide covers the most important parameters across hhblits, hhsearch, and hhalign.

Core Parameters

Input/Output

-i

string

required

Input file containing query sequence or alignment in FASTA, A2M, or A3M format

hhblits -i query.fasta -d database -o results.hhr

-o

string

required

Output file for search results in HHR format

hhsearch -i query.a3m -d pdb70 -o output.hhr

-d

string

required

Database basename (without file extensions)

hhblits -i query.fas -d /data/uniclust30 -o results.hhr

Alignment Output Formats

-oa3m

string

Output alignment in A3M format (MSA with query)

hhblits -i query.fas -d database -oa3m result.a3m

-opsi

string

Output alignment in PSI-BLAST format

-ohhm

string

Output HMM file

-blasttab

string

Output in BLAST tab format (M8)

-atab

string

Output alignment in tabular format with residue indices

Search Control

E-value and Probability Thresholds

# Include hits with E-value <= 0.001 in hit list
hhsearch -i query.a3m -d pdb70 -E 0.001 -o results.hhr

# Include hits with E-value <= 0.01 in output alignment
hhsearch -i query.a3m -d pdb70 -e 0.01 -oa3m output.a3m

-E

float

default:"1E+03"

Maximum E-value threshold for inclusion in hit list

-e

float

default:"1E-03"

Maximum E-value threshold for inclusion in output alignment and HMM

-p

float

default:"20"

Minimum probability threshold (0-100) for inclusion in hit list

Number of Hits

-Z

int

default:"500"

Maximum number of lines in hit list

-z

int

default:"10"

Minimum number of lines in hit list

-B

int

default:"500"

Maximum number of alignments in alignment list

-b

int

default:"10"

Minimum number of alignments in alignment list

Gap Penalties

Gap Opening and Extension

Gap penalties are specified as positive numbers but applied as penalties (negative scores).

-gapd

float

default:"0.15"

Gap open penalty factor for deletions (query gaps)

-gape

float

default:"1.0"

Gap extension penalty factor

-gapf

float

default:"0.6"

Factor for increasing gap opening penalty for deletions

-gapg

float

default:"0.6"

Factor for increasing gap opening penalty for insertions (template gaps)

-gaph

float

default:"0.6"

Factor for increasing gap extension penalty for deletions

-gapi

float

default:"0.6"

Factor for increasing gap extension penalty for insertions

End Gap Penalties

-egq

float

default:"0.0"

Penalty for end gaps when query not fully covered

-egt

float

default:"0.0"

Penalty for end gaps when template not fully covered

Pseudocounts

Pseudocounts prevent overfitting to the query alignment.

Admixture Parameters

-pc_hhm_context_engine.admix

int

default:"2"

Admixture mode:

1: Constant admixture
2: HHsearch admixture (default)
3: CSBlast admixture

-pcm

int

default:"0"

Pseudocount mode for HMM:

0: Default (context-dependent if available)
1: Constant admixture
2: HHsearch admixture
3: CSBlast admixture

-pca

float

default:"1.0"

Pseudocount admixture parameter a

-pcb

float

default:"1.5"

Pseudocount admixture parameter b

-pcc

float

default:"1.0"

Pseudocount admixture parameter c (HHsearch mode only)

Context-Specific Pseudocounts

-csb

float

default:"1.7"

Context-specific pseudocount parameter b

-csw

float

default:"1.7"

Context-specific pseudocount weight

-cs

string

Path to context library for context-specific pseudocounts

Secondary Structure Scoring

-ssm

int

default:"2"

Secondary structure scoring mode:

0: No SS scoring
1: SS scoring after alignment
2: SS scoring during alignment (default)

-ssw

float

default:"0.11"

Weight of secondary structure score

-ssa

float

default:"1.0"

SS state evolution matrix: M1 = (1-ssa)I + ssaM0

Alignment Mode

Local Alignment
Global Alignment

Local (default)

# Align best matching regions
hhsearch -i query.a3m -d pdb70 -loc -o results.hhr

Local alignment (Smith-Waterman) finds the best matching subsequences.

Global

# Align entire sequences
hhsearch -i query.a3m -d pdb70 -glob -o results.hhr

Global alignment (Needleman-Wunsch) aligns sequences from end to end.

-loc

flag

Local alignment mode (default)

-glob

flag

Global alignment mode

Realignment and MAC Algorithm

-realign

flag

Realign hits using MAC (Maximum Accuracy) algorithm

-mact

float

default:"0.35"

MAC alignment threshold (controls greediness at alignment ends)

Lower values (e.g., 0.1) produce more conservative alignments

-realign_max

int

default:"500"

Maximum number of hits to realign

-alt

int

default:"1"

Number of alternative alignments to find

Filtering Parameters

Query MSA Filtering

-id

int

default:"90"

Maximum sequence identity with all other sequences (0-100)

-qid

int

default:"0"

Minimum sequence identity with query sequence (0-100)

-qsc

float

default:"-20.0"

Minimum score per column with query sequence

-cov

int

default:"0"

Minimum coverage threshold (0-100)

-diff

int

default:"0"

Filter most different sequences that pass other thresholds

Match State Assignment

-M

string

default:"a2m"

Match state assignment:

a2m: By upper/lower case (default)
first: By first sequence
[0-100]: By percentage rule (e.g., -M 50)

-Mgaps

int

default:"100"

Maximum percentage of gaps in match states

HHblits-Specific Parameters

Iterative Search

-n

int

default:"2"

Number of search iterations

Three iterations

hhblits -i query.fas -d uniclust30 -n 3 -oa3m result.a3m

-e

float

default:"1E-03"

E-value threshold for inclusion in next iteration

Prefiltering

-prefilter

flag

default:"enabled"

Use prefiltering to speed up search (enabled by default)

-noprefilt

flag

Disable prefiltering

-prefilter_evalue_thresh

float

default:"1000"

E-value threshold for prefilter

-min_prefilter_hits

int

default:"100"

Minimum number of hits to pass prefilter

Diversity Control

-neffmax

float

default:"20"

Maximum Neff (diversity) for MSA after iteration

Performance Parameters

Multi-threading

-cpu

int

default:"2"

Number of CPU threads to use

Using 8 threads

hhblits -i query.fas -d database -cpu 8 -o results.hhr

Memory Management

-maxres

int

default:"15000"

Maximum number of HMM states (match states)

-maxseq

int

default:"65535"

Maximum number of sequences in input alignment

-maxmem

float

default:"3.0"

Maximum memory (GB) for realignment

Output Customization

Alignment Display

-seq

int

default:"1"

Maximum number of query/template sequences to show in alignment

-aliw

int

default:"80"

Alignment width (characters per line)

-showcons

flag

Show consensus sequence in alignment

-showdssp

flag

Show DSSP secondary structure in alignment

-showpred

flag

Show predicted secondary structure in alignment

Verbosity

-v

int

default:"2"

Verbosity level:

0: Errors only
1: Warnings
2: Info (default)
3: Debug

Advanced Scoring Parameters

Score Offset and Correlation

-shift

float

default:"-0.03"

Score offset for match-match states

Adjusts the scoring to balance sensitivity and specificity

-corr

float

default:"0.1"

Weight of correlations between scores with |i-j| <= 4

-sc

int

default:"1"

Column score mode:

0: No amino acid composition correction
1: 1/2(query_avg + template_avg) (default)
2: Template average frequencies
3: Query average frequencies

Minimum Overlap

-min-overlap

int

default:"0"

Minimum overlap between query and template

Require 30-residue overlap

hhsearch -i query.a3m -d pdb70 -min-overlap 30 -o results.hhr

Practical Examples

# High sensitivity, more iterations
hhblits -i query.fas -d uniclust30 \
  -n 3 \
  -e 1E-5 \
  -realign \
  -mact 0.1 \
  -cpu 8 \
  -o results.hhr \
  -oa3m output.a3m

Parameter Files

Save frequently used parameters in a file:

hhblits.params

-n 3
-e 1E-5
-cpu 8
-realign
-mact 0.1

Use with @ syntax:

hhblits -i query.fas -d database @hhblits.params -o results.hhr

Getting Started

Core Tools

Utility Tools

Guides

Advanced

​Overview

​Core Parameters

​Input/Output

​Alignment Output Formats

​Search Control

​E-value and Probability Thresholds

​Number of Hits

​Gap Penalties

​Gap Opening and Extension

​End Gap Penalties

​Pseudocounts

​Admixture Parameters

​Context-Specific Pseudocounts

​Secondary Structure Scoring

​Alignment Mode

​Realignment and MAC Algorithm

​Filtering Parameters

​Query MSA Filtering

​Match State Assignment

​HHblits-Specific Parameters

​Iterative Search

​Prefiltering

​Diversity Control

​Performance Parameters

​Multi-threading

​Memory Management

​Output Customization

​Alignment Display

​Verbosity

​Advanced Scoring Parameters

​Score Offset and Correlation

​Minimum Overlap

​Practical Examples

​Parameter Files

​Next Steps

Algorithms

Compilation

Build docs developers (and LLMs) love

Overview

Core Parameters

Input/Output

Alignment Output Formats

Search Control

E-value and Probability Thresholds

Number of Hits

Gap Penalties

Gap Opening and Extension

End Gap Penalties

Pseudocounts

Admixture Parameters

Context-Specific Pseudocounts

Secondary Structure Scoring

Alignment Mode

Realignment and MAC Algorithm

Filtering Parameters

Query MSA Filtering

Match State Assignment

HHblits-Specific Parameters

Iterative Search

Prefiltering

Diversity Control

Performance Parameters

Multi-threading

Memory Management

Output Customization

Alignment Display

Verbosity

Advanced Scoring Parameters

Score Offset and Correlation

Minimum Overlap

Practical Examples

Parameter Files

Next Steps