API reference

The nucleotide sequence to be processed. ASCII characters other than 'A', 'C', 'G', 'T', 'a', 'c', 'g', and 't' are considered ambiguous bases. Non-ASCII characters are rejected. The minimum allowed sequence length is 4 bases.

The length of the window used by symmetric DUST algorithm. The minimum allowed value is 4.

Score threshold for identifying low-complexity regions. Higher values result in fewer regions being masked.

The nucleotide sequence that was provided as input.

The length of the window used by symmetric DUST algorithm.

Score threshold for identifying low-complexity regions.

A tuple of tuples representing the start and end positions of the low-complexity regions identified in the sequence.

The total number of bases that were masked.

Returns the sequence with low-complexity regions masked.

If the input parameters violate the following constraints:

• sequence contains a non-ASCII character
• sequence length < 4
• window_size < 4

If the input parameters are not of the expected type.

If window_size or score_threshold cannot be represented by its unsigned integer type, including negative or excessively large values.

Nucleotide sequence to process. ASCII characters other than 'A', 'C', 'G', 'T', 'a', 'c', 'g', and 't' are treated as ambiguous bases. Non-ASCII characters are rejected. Must contain at least kmer + 1 bases.

Maximum sliding-window size. Larger windows can detect longer repeats but use more memory. Must be in the range [kmer + 1, 65535].

Score threshold for identifying low-complexity regions. Higher values result in fewer regions being masked. Must be finite and greater than 0.0.

The k-mer length used by the Longdust algorithm. Must be in the range [1, 12].

GC content for bias correction. If None (default), assume a uniform base composition. If 'auto', compute the GC content from the unambiguous bases ('A', 'C', 'G', and 'T', case-insensitive) of the input sequence, or use 0.0 if there are none. If a float, it must be between 0.0 and 1.0.

Maximum allowable score drop for X-drop extension termination. During backward scanning, extension continues while (max_score - current_score) does not exceed (score_threshold * xdrop). Once the score drops by more than this amount from the peak score observed during the scan, extension stops immediately. Lower values enforce stricter extensions and tighter boundaries, potentially truncating part of the low-complexity region, whereas higher values allow more permissive extensions and looser boundaries, which may include non-low-complexity regions. If set to None, X-drop is disabled.

Minimum k-mer frequency to start a backward scan. Must be in [2, 65535]. Lower values are more sensitive but slower.

If True, use approximate mode for guaranteed O(L*w) time complexity. In this mode, only the first candidate starting position is examined during backward scanning, rather than checking all candidates to find the optimal one.

If True, only process the forward strand. Otherwise, both strands are processed.

The nucleotide sequence that was provided as input.

The size of the sliding window used to scan for low-complexity regions.

Score threshold for determining low-complexity regions.

k-mer length.

Option used for GC bias correction. Can be None (a uniform base composition was assumed), 'auto' (GC was computed from the unambiguous bases of the input sequence), or a float between 0.0 and 1.0 (provided by the user).

Multiplier used to derive the X-drop extension threshold.

Minimum k-mer frequency to trigger backward scan.

Whether approximate mode was enabled.

Whether only the forward strand was processed.

A tuple of tuples representing the start and end positions of the low-complexity regions identified in the sequence.

The total number of bases that were masked.

Returns the sequence with low-complexity regions masked.

If the input parameters violate the following constraints:

• sequence contains a non-ASCII character
• sequence length < kmer + 1
• window_size is not in [kmer + 1, 65535]
• kmer is not in [1, 12]
• score_threshold is non-finite or not greater than 0.0
• min_start_cnt < 2
• xdrop is 0
• gc is invalid (not 'auto', None, or float between 0.0 and 1.0)

If the input parameters are not of the expected type.

If an integer cannot be represented by its parameter type, including a negative value or min_start_cnt above 65535.

The nucleotide or protein sequence to be processed. When protein is False, ASCII characters other than 'A', 'C', 'G', and 'T' (case-insensitive) are treated as ambiguous residues. When protein is True, ASCII characters other than the standard 20 amino acids (case-insensitive) are considered ambiguous. Non-ASCII characters are rejected in either mode. Must contain at least one character.

Determines whether the input is treated as a protein sequence (True) or as a nucleotide sequence (False).

Probability of transitioning from the background state to a repeat state. Must be in the range [0.0, 1.0).

Probability of ending a repeat (transitioning from a repeat state back to the background). Must be in the range [0.0, 1.0].

Probability decay from one period offset to the next. Must be in the range (0.0, 1.0] and be a normal finite number.

Maximum repeat period (cycle length) to consider. If None (default), it resolves to 50 for protein sequences and 100 for nucleotide sequences. Must be in the range [1, 2147483647].

Cost of opening a gap within a repeat.

Cost of extending a gap by one more letter. If None (default), gaps within repeats are disabled. If set, must be greater than 0.

Posterior probability threshold at or above which a position is considered part of a repeat. Used by intervals and mask. Has no effect on probabilities or repeat_units. Must be in the range [0.0, 1.0].

Minimum copy-number estimate from the Viterbi path for a tandem repeat tract to be reported by repeat_units. Selecting a period does not recompute this estimate from the selected unit. Must be finite and non-negative. Has no effect on intervals, mask, and probabilities.

The sequence that was provided as input.

Whether the sequence was treated as a protein sequence.

Probability of a repeat starting per position.

Probability of a repeat ending per position.

Probability decay per unit increase in repeat period.

Maximum tandem repeat period (in letters) considered. Resolved from None to 50 (protein) or 100 (nucleotide) when not provided explicitly.

Cost of opening a gap within a repeat.

Cost of extending a gap by one more letter, or None if gaps within repeats are disabled.

Posterior probability threshold at or above which a position is considered part of a repeat by intervals and mask. Has no effect on probabilities or repeat_units.

Minimum copy number used to filter the tracts returned by repeat_units.

An immutable tuple of tuples representing the start and end positions of the tandem repeat regions identified in the sequence, based on per-position posterior probabilities.

The total number of bases/residues that were masked.

The per-position posterior probability of being part of a tandem repeat, one value per character of sequence, in order. Computed once at construction time.

Returns the sequence with tandem repeat regions masked.

Returns the consensus tandem repeat unit(s), optionally for a selected period, identified via a Viterbi decode independent of intervals.

If the input parameters violate the following constraints:

• sequence is empty
• sequence contains a non-ASCII character
• repeat_start is not in [0.0, 1.0)
• repeat_end is not in [0.0, 1.0]
• decay is not a normal finite number in (0.0, 1.0]
• score_threshold is not in [0.0, 1.0]
• max_period is not in [1, 2147483647]
• gap_extend is 0
• the combined repeat/gap probabilities form an invalid distribution (i.e. repeat_end + 2 * first_gap_prob > 1.0, where the gap probabilities are derived from gap_open and gap_extend). Raising gap_open or gap_extend lowers first_gap_prob, so this is resolved by increasing them or by lowering repeat_end.
• min_copy_number is not finite, or is negative

If the input parameters are not of the expected type.

If max_period, gap_open, or gap_extend cannot be represented by its unsigned integer type, including negative or excessively large values.