Source code for pydna.design

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
# Copyright 2013-2023 by Björn Johansson.  All rights reserved.
# This code is part of the Python-dna distribution and governed by its
# license.  Please see the LICENSE.txt file that should have been included
# as part of this package.
"""This module contain functions for primer design for various purposes.

- :func:primer_design for designing primers for a sequence or a matching primer for an existing primer. Returns an :class:`Amplicon` object (same as the :mod:`amplify` module returns).

- :func:assembly_fragments Adds tails to primers for a linear assembly through homologous recombination or Gibson assembly.

- :func:circular_assembly_fragments Adds tails to primers for a circular assembly through homologous recombination or Gibson assembly.

"""

from pydna.tm import tm_default
import math
import copy
from pydna.amplicon import Amplicon
from pydna.amplify import Anneal
from pydna.amplify import pcr
from pydna.dseqrecord import Dseqrecord
from pydna.primer import Primer
import operator
from typing import Tuple
from itertools import pairwise, product
import re


def _design_primer(
    target_tm: float,
    template: Dseqrecord,
    limit: int,
    tm_func,
    starting_length: int = 0,
) -> Tuple[float, str]:
    """returns a tuple (temp, primer)"""

    if starting_length < limit:
        starting_length = limit

    length = starting_length
    tlen = len(template)

    tmp = max(0, tm_func(str(template.seq[:length])))

    p = str(template.seq[:length])

    if tmp < target_tm:
        condition = operator.le
        increment = 1
    else:
        condition = operator.ge
        increment = -1
    while condition(tmp, target_tm):
        prev_temp = tmp
        prev_primer = p
        length += increment
        p = str(template.seq[:length])
        tmp = tm_func(p)
        if length >= tlen:
            break
        # Never go below the limit
        if length < limit:
            return template.seq[:limit]

    if abs(target_tm - tmp) < abs(target_tm - prev_temp):
        return p
    else:
        return prev_primer


[docs] def primer_design( template, fp=None, rp=None, limit=13, target_tm=55.0, tm_func=tm_default, estimate_function=None, **kwargs, ): """This function designs a forward primer and a reverse primer for PCR amplification of a given template sequence. The template argument is a Dseqrecord object or equivalent containing the template sequence. The optional fp and rp arguments can contain an existing primer for the sequence (either the forward or reverse primer). One or the other primers can be specified, not both (since then there is nothing to design!, use the pydna.amplify.pcr function instead). The limit argument is the minimum length of the primer. The default value is 13. If one of the primers is given, the other primer is designed to match in terms of Tm. If both primers are designed, they will be designed to target_tm tm_func is a function that takes an ascii string representing an oligonuceotide as argument and returns a float. Some useful functions can be found in the :mod:`pydna.tm` module, but can be substituted for a custom made function. estimate_function is a tm_func-like function that is used to get a first guess for the primer design, that is then used as starting point for the final result. This is useful when the tm_func function is slow to calculate (e.g. it relies on an external API, such as the NEB primer design API). The estimate_function should be faster than the tm_func function. The default value is `None`. To use the default `tm_func` as estimate function to get the NEB Tm faster, you can do: `primer_design(dseqr, target_tm=55, tm_func=tm_neb, estimate_function=tm_default)`. The function returns a pydna.amplicon.Amplicon class instance. This object has the object.forward_primer and object.reverse_primer properties which contain the designed primers. Parameters ---------- template : pydna.dseqrecord.Dseqrecord a Dseqrecord object. The only required argument. fp, rp : pydna.primer.Primer, optional optional pydna.primer.Primer objects containing one primer each. target_tm : float, optional target tm for the primers, set to 55°C by default. tm_func : function Function used for tm calculation. This function takes an ascii string representing an oligonuceotide as argument and returns a float. Some useful functions can be found in the :mod:`pydna.tm` module, but can be substituted for a custom made function. Returns ------- result : Amplicon Examples -------- >>> from pydna.dseqrecord import Dseqrecord >>> t=Dseqrecord("atgactgctaacccttccttggtgttgaacaagatcgacgacatttcgttcgaaacttacgatg") >>> t Dseqrecord(-64) >>> from pydna.design import primer_design >>> ampl = primer_design(t) >>> ampl Amplicon(64) >>> ampl.forward_primer f64 17-mer:5'-atgactgctaacccttc-3' >>> ampl.reverse_primer r64 18-mer:5'-catcgtaagtttcgaacg-3' >>> print(ampl.figure()) 5atgactgctaacccttc...cgttcgaaacttacgatg3 |||||||||||||||||| 3gcaagctttgaatgctac5 5atgactgctaacccttc3 ||||||||||||||||| 3tactgacgattgggaag...gcaagctttgaatgctac5 >>> pf = "GGATCC" + ampl.forward_primer >>> pr = "GGATCC" + ampl.reverse_primer >>> pf f64 23-mer:5'-GGATCCatgactgct..ttc-3' >>> pr r64 24-mer:5'-GGATCCcatcgtaag..acg-3' >>> from pydna.amplify import pcr >>> pcr_prod = pcr(pf, pr, t) >>> print(pcr_prod.figure()) 5atgactgctaacccttc...cgttcgaaacttacgatg3 |||||||||||||||||| 3gcaagctttgaatgctacCCTAGG5 5GGATCCatgactgctaacccttc3 ||||||||||||||||| 3tactgacgattgggaag...gcaagctttgaatgctac5 >>> print(pcr_prod.seq) GGATCCatgactgctaacccttccttggtgttgaacaagatcgacgacatttcgttcgaaacttacgatgGGATCC >>> from pydna.primer import Primer >>> pf = Primer("atgactgctaacccttccttggtgttg", id="myprimer") >>> ampl = primer_design(t, fp = pf) >>> ampl.forward_primer myprimer 27-mer:5'-atgactgctaaccct..ttg-3' >>> ampl.reverse_primer r64 32-mer:5'-catcgtaagtttcga..atc-3' """ def design(target_tm, template): if estimate_function: first_guess = _design_primer(target_tm, template, limit, estimate_function) return _design_primer(target_tm, template, limit, tm_func, len(first_guess)) else: return _design_primer(target_tm, template, limit, tm_func) if not fp and not rp: fp = Primer((design(target_tm, template))) target_tm = tm_func(str(fp.seq)) rp = Primer(design(target_tm, template.reverse_complement())) elif fp and not rp: try: fp = Anneal((fp,), template).forward_primers.pop() except IndexError: raise ValueError("Forward primer does not anneal") except Exception: # pragma: no cover print("Unexpected error") raise target_tm = tm_func(fp.footprint) rp = Primer(design(target_tm, template.reverse_complement())) elif not fp and rp: try: rp = Anneal((rp,), template).reverse_primers.pop() except IndexError: raise ValueError("Reverse primer does not anneal") except Exception: # pragma: no cover print("Unexpected error") raise target_tm = tm_func(rp.footprint) fp = Primer(design(target_tm, template)) else: raise ValueError("Specify maximum one of the two primers.") if fp.id == "id": # <unknown id> fp.id = "f{}".format(len(template)) if rp.id == "id": rp.id = "r{}".format(len(template)) if fp.name == "name": fp.name = "f{}".format(len(template)) if rp.name == "name": rp.name = "r{}".format(len(template)) fp.description = fp.id + " " + template.accession rp.description = rp.id + " " + template.accession ampl = Anneal((fp, rp), template, limit=limit) prod = ampl.products[0] if ampl.products else Amplicon("") if len(ampl.products) > 1: import warnings from pydna import _PydnaWarning warnings.warn( "designed primers do not yield a unique PCR product", _PydnaWarning ) return prod
[docs] def assembly_fragments(f, overlap=35, maxlink=40, circular=False): """This function return a list of :mod:`pydna.amplicon.Amplicon` objects where primers have been modified with tails so that the fragments can be fused in the order they appear in the list by for example Gibson assembly or homologous recombination. Given that we have two linear :mod:`pydna.amplicon.Amplicon` objects a and b we can modify the reverse primer of a and forward primer of b with tails to allow fusion by fusion PCR, Gibson assembly or in-vivo homologous recombination. The basic requirements for the primers for the three techniques are the same. :: _________ a _________ / \\ agcctatcatcttggtctctgca ||||| <gacgt agcct> ||||| tcggatagtagaaccagagacgt __________ b ________ / \\ TTTATATCGCATGACTCTTCTTT ||||| <AGAAA TTTAT> ||||| AAATATAGCGTACTGAGAAGAAA agcctatcatcttggtctctgcaTTTATATCGCATGACTCTTCTTT |||||||||||||||||||||||||||||||||||||||||||||| tcggatagtagaaccagagacgtAAATATAGCGTACTGAGAAGAAA \\___________________ c ______________________/ Design tailed primers incorporating a part of the next or previous fragment to be assembled. :: agcctatcatcttggtctctgca ||||||||||||||||||||||| gagacgtAAATATA ||||||||||||||||||||||| tcggatagtagaaccagagacgt TTTATATCGCATGACTCTTCTTT ||||||||||||||||||||||| ctctgcaTTTATAT ||||||||||||||||||||||| AAATATAGCGTACTGAGAAGAAA PCR products with flanking sequences are formed in the PCR process. :: agcctatcatcttggtctctgcaTTTATAT |||||||||||||||||||||||||||||| tcggatagtagaaccagagacgtAAATATA \\____________/ identical sequences ____________ / \\ ctctgcaTTTATATCGCATGACTCTTCTTT |||||||||||||||||||||||||||||| gagacgtAAATATAGCGTACTGAGAAGAAA The fragments can be fused by any of the techniques mentioned earlier to form c: :: agcctatcatcttggtctctgcaTTTATATCGCATGACTCTTCTTT |||||||||||||||||||||||||||||||||||||||||||||| tcggatagtagaaccagagacgtAAATATAGCGTACTGAGAAGAAA The first argument of this function is a list of sequence objects containing Amplicons and other similar objects. **At least every second sequence object needs to be an Amplicon** This rule exists because if a sequence object is that is not a PCR product is to be fused with another fragment, that other fragment needs to be an Amplicon so that the primer of the other object can be modified to include the whole stretch of sequence homology needed for the fusion. See the example below where a is a non-amplicon (a linear plasmid vector for instance) :: _________ a _________ __________ b ________ / \\ / \\ agcctatcatcttggtctctgca <--> TTTATATCGCATGACTCTTCTTT ||||||||||||||||||||||| ||||||||||||||||||||||| tcggatagtagaaccagagacgt <AGAAA TTTAT> ||||||||||||||||||||||| <--> AAATATAGCGTACTGAGAAGAAA agcctatcatcttggtctctgcaTTTATATCGCATGACTCTTCTTT |||||||||||||||||||||||||||||||||||||||||||||| tcggatagtagaaccagagacgtAAATATAGCGTACTGAGAAGAAA \\___________________ c ______________________/ In this case only the forward primer of b is fitted with a tail with a part a: :: agcctatcatcttggtctctgca ||||||||||||||||||||||| tcggatagtagaaccagagacgt TTTATATCGCATGACTCTTCTTT ||||||||||||||||||||||| <AGAAA tcttggtctctgcaTTTATAT ||||||||||||||||||||||| AAATATAGCGTACTGAGAAGAAA PCR products with flanking sequences are formed in the PCR process. :: agcctatcatcttggtctctgcaTTTATAT |||||||||||||||||||||||||||||| tcggatagtagaaccagagacgtAAATATA \\____________/ identical sequences ____________ / \\ ctctgcaTTTATATCGCATGACTCTTCTTT |||||||||||||||||||||||||||||| gagacgtAAATATAGCGTACTGAGAAGAAA The fragments can be fused by for example Gibson assembly: :: agcctatcatcttggtctctgcaTTTATAT |||||||||||||||||||||||||||||| tcggatagtagaacca TCGCATGACTCTTCTTT |||||||||||||||||||||||||||||| gagacgtAAATATAGCGTACTGAGAAGAAA to form c: :: agcctatcatcttggtctctgcaTTTATATCGCATGACTCTTCTTT |||||||||||||||||||||||||||||||||||||||||||||| tcggatagtagaaccagagacgtAAATATAGCGTACTGAGAAGAAA The first argument of this function is a list of sequence objects containing Amplicons and other similar objects. The overlap argument controls how many base pairs of overlap required between adjacent sequence fragments. In the junction between Amplicons, tails with the length of about half of this value is added to the two primers closest to the junction. :: > < Amplicon1 Amplicon2 > < > <- Amplicon1 Amplicon2 -> < In the case of an Amplicon adjacent to a Dseqrecord object, the tail will be twice as long (1*overlap) since the recombining sequence is present entirely on this primer: :: Dseqrecd1 Amplicon1 > < Dseqrecd1 Amplicon1 --> < Note that if the sequence of DNA fragments starts or stops with an Amplicon, the very first and very last prinmer will not be modified i.e. assembles are always assumed to be linear. There are simple tricks around that for circular assemblies depicted in the last two examples below. The maxlink arguments controls the cut off length for sequences that will be synhtesized by adding them to primers for the adjacent fragment(s). The argument list may contain short spacers (such as spacers between fusion proteins). :: Example 1: Linear assembly of PCR products (pydna.amplicon.Amplicon class objects) ------ > < > < Amplicon1 Amplicon3 Amplicon2 Amplicon4 > < > < pydna.design.assembly_fragments > <- -> <- pydna.assembly.Assembly Amplicon1 Amplicon3 Amplicon2 Amplicon4 ➤ Amplicon1Amplicon2Amplicon3Amplicon4 -> <- -> < Example 2: Linear assembly of alternating Amplicons and other fragments > < > < Amplicon1 Amplicon2 Dseqrecd1 Dseqrecd2 pydna.design.assembly_fragments > <-- --> <-- pydna.assembly.Assembly Amplicon1 Amplicon2 Dseqrecd1 Dseqrecd2 ➤ Amplicon1Dseqrecd1Amplicon2Dseqrecd2 Example 3: Linear assembly of alternating Amplicons and other fragments Dseqrecd1 Dseqrecd2 Amplicon1 Amplicon2 > < --> < pydna.design.assembly_fragments pydna.assembly.Assembly Dseqrecd1 Dseqrecd2 Amplicon1 Amplicon2 ➤ Dseqrecd1Amplicon1Dseqrecd2Amplicon2 --> <-- --> < Example 4: Circular assembly of alternating Amplicons and other fragments -> <== Dseqrecd1 Amplicon2 Amplicon1 Dseqrecd1 --> <- pydna.design.assembly_fragments pydna.assembly.Assembly -> <== Dseqrecd1 Amplicon2 -Dseqrecd1Amplicon1Amplicon2- Amplicon1 ➤ | | --> <- ----------------------------- ------ Example 5: Circular assembly of Amplicons > < > < Amplicon1 Amplicon3 Amplicon2 Amplicon1 > < > < pydna.design.assembly_fragments > <= -> <- Amplicon1 Amplicon3 Amplicon2 Amplicon1 -> <- +> < make new Amplicon using the Amplicon1.template and the last fwd primer and the first rev primer. pydna.assembly.Assembly +> <= -> <- Amplicon1 Amplicon3 -Amplicon1Amplicon2Amplicon3- Amplicon2 ➤ | | -> <- ----------------------------- Parameters ---------- f : list of :mod:`pydna.amplicon.Amplicon` and other Dseqrecord like objects list Amplicon and Dseqrecord object for which fusion primers should be constructed. overlap : int, optional Length of required overlap between fragments. maxlink : int, optional Maximum length of spacer sequences that may be present in f. These will be included in tails for designed primers. circular : bool, optional If True, the assembly is circular. If False, the assembly is linear. Returns ------- seqs : list of :mod:`pydna.amplicon.Amplicon` and other Dseqrecord like objects :mod:`pydna.amplicon.Amplicon` objects :: [Amplicon1, Amplicon2, ...] Examples -------- >>> from pydna.dseqrecord import Dseqrecord >>> from pydna.design import primer_design >>> a=primer_design(Dseqrecord("atgactgctaacccttccttggtgttgaacaagatcgacgacatttcgttcgaaacttacgatg")) >>> b=primer_design(Dseqrecord("ccaaacccaccaggtaccttatgtaagtacttcaagtcgccagaagacttcttggtcaagttgcc")) >>> c=primer_design(Dseqrecord("tgtactggtgctgaaccttgtatcaagttgggtgttgacgccattgccccaggtggtcgtttcgtt")) >>> from pydna.design import assembly_fragments >>> # We would like a circular recombination, so the first sequence has to be repeated >>> fa1,fb,fc,fa2 = assembly_fragments([a,b,c,a]) >>> # Since all fragments are Amplicons, we need to extract the rp of the 1st and fp of the last fragments. >>> from pydna.amplify import pcr >>> fa = pcr(fa2.forward_primer, fa1.reverse_primer, a) >>> [fa,fb,fc] [Amplicon(100), Amplicon(101), Amplicon(102)] >>> fa.name, fb.name, fc.name = "fa fb fc".split() >>> from pydna.assembly import Assembly >>> assemblyobj = Assembly([fa,fb,fc]) >>> assemblyobj Assembly fragments....: 100bp 101bp 102bp limit(bp)....: 25 G.nodes......: 6 algorithm....: common_sub_strings >>> assemblyobj.assemble_linear() [Contig(-231), Contig(-166), Contig(-36)] >>> assemblyobj.assemble_circular()[0].seguid() 'cdseguid=85t6tfcvWav0wnXEIb-lkUtrl4s' >>> (a+b+c).looped().seguid() 'cdseguid=85t6tfcvWav0wnXEIb-lkUtrl4s' >>> print(assemblyobj.assemble_circular()[0].figure()) -|fa|36 | \\/ | /\\ | 36|fb|36 | \\/ | /\\ | 36|fc|36 | \\/ | /\\ | 36- | | -------------------- >>> """ # Recursive call for circular assemblies if circular: fragments = assembly_fragments( f + f[0:1], overlap=overlap, maxlink=maxlink, circular=False ) if hasattr(fragments[0], "template"): fragments[0] = pcr( (fragments[-1].forward_primer, fragments[0].reverse_primer), fragments[0].template, ) return fragments[:-1] # sanity check for arguments nf = [item for item in f if len(item) > maxlink] if not all( hasattr(i[0], "template") or hasattr(i[1], "template") for i in zip(nf, nf[1:]) ): raise ValueError( "Every second fragment larger than maxlink has to be an Amplicon object" ) f = [copy.copy(f) for f in f] first_fragment_length = len(f[0]) last_fragment_length = len(f[-1]) if first_fragment_length <= maxlink: # first fragment should be removed and added to second fragment (new first fragment) forward primer f[1].forward_primer = f[0].seq._data.decode("ASCII") + f[1].forward_primer f = f[1:] # else: if last_fragment_length <= maxlink: f[-2].reverse_primer = ( f[-1].seq.reverse_complement()._data.decode("ASCII") + f[-2].reverse_primer ) f = f[:-1] # else: empty = Dseqrecord("") tail_length = math.ceil(overlap / 2) for i in range(len(f) - 1): first = f[i] secnd = f[i + 1] secnd_len = len(secnd) if secnd_len <= maxlink: third = f[i + 2] if hasattr(f[i], "template") and hasattr(third, "template"): # "secnd is is flanked by amplicons, so half of secnd should be added each flanking primers" # ) first.reverse_primer = ( secnd.seq.reverse_complement()._data.decode("ASCII")[ secnd_len // 2 : ] + first.reverse_primer ) third.forward_primer = ( secnd.seq._data.decode("ASCII")[secnd_len // 2 :] + third.forward_primer ) lnk = ( third.seq.reverse_complement()._data.decode("ASCII") + secnd.reverse_complement().seq._data.decode("ASCII")[ : secnd_len // 2 ] )[-tail_length:] first.reverse_primer = lnk + first.reverse_primer lnk = ( first.seq._data.decode("ASCII") + secnd.seq._data.decode("ASCII")[: secnd_len // 2] )[-tail_length:] third.forward_primer = lnk + third.forward_primer elif hasattr(first, "template"): first.reverse_primer = ( secnd.seq.reverse_complement()._data.decode("ASCII") + first.reverse_primer ) lnk = str(third.seq[:overlap].reverse_complement()) first.reverse_primer = lnk + first.reverse_primer elif hasattr(third, "template"): third.forward_primer = ( secnd.seq._data.decode("ASCII") + third.forward_primer ) lnk = str(first.seq[-overlap:]) third.forward_primer = lnk + third.forward_primer secnd = empty f[i + 2] = third else: # secnd is larger than maxlink if hasattr(first, "template") and hasattr(secnd, "template"): lnk = str(first.seq[-tail_length:]) # #_module_logger.debug("4 %s", lnk) secnd.forward_primer = lnk + secnd.forward_primer lnk = str(secnd.seq[:tail_length].reverse_complement()) # #_module_logger.debug("5 %s", lnk) first.reverse_primer = lnk + first.reverse_primer elif hasattr(first, "template"): lnk = str(secnd.seq[:overlap].reverse_complement()) # #_module_logger.debug("6 %s", lnk) first.reverse_primer = lnk + first.reverse_primer elif hasattr(secnd, "template"): lnk = str(first.seq[-overlap:]) # #_module_logger.debug("7 %s", lnk) secnd.forward_primer = lnk + secnd.forward_primer f[i] = first f[i + 1] = secnd f = [item for item in f if len(item)] return [ ( pcr( p.forward_primer, p.reverse_primer, p.template, limit=min((p.forward_primer._fp, p.reverse_primer._fp)), ) if hasattr(p, "template") else p ) for p in f ]
[docs] def circular_assembly_fragments(f, overlap=35, maxlink=40): """ Equivalent to `assembly_fragments` with `circular=True`. Deprecated, kept for backward compatibility. Use `assembly_fragments` with `circular=True` instead. """ import warnings warnings.warn( "The circular_assembly_fragments function is deprecated. Use assembly_fragments with circular=True instead.", DeprecationWarning, stacklevel=2, ) return assembly_fragments(f, overlap=overlap, maxlink=maxlink, circular=True)
[docs] def user_assembly_design( f: list[Amplicon], max_overlap: int = 15, min_overlap: int = 4, max_tail=50 ) -> list[Amplicon]: import warnings warnings.warn( "The user_assembly_design function is experimental and " "may change in future versions.", category=FutureWarning, stacklevel=2, ) assert max_overlap > min_overlap, ( f"max_overlap ({max_overlap}) " "has to be larger than min_overlap " f"({min_overlap})" ) amplicons = [] for fragment in f: amplicons.append(primer_design(fragment)) flag = True for ths, nxt in pairwise(amplicons): A_positions_in_ths = [m.start() for m in re.finditer("A|a", str(ths.seq))] T_positions_in_nxt = [m.start() for m in re.finditer("T|t", str(nxt.seq))] for ths_a, ths_t in zip(A_positions_in_ths[::-1], T_positions_in_nxt): sticky_length = ths_t + len(ths) - ths_a if sticky_length < min_overlap: continue if sticky_length > max_overlap: flag = False break rp = bytearray( nxt.seq[: ths_t + 1].rc()._data + ths.reverse_primer.seq._data ) fp = bytearray(ths.seq[ths_a:]._data + nxt.forward_primer.seq._data) fp[sticky_length] = ord(b"U") rp[sticky_length] = ord(b"U") ths.reverse_primer = Primer(rp) nxt.forward_primer = Primer(fp) break # Primers were designed. else: flag = False if flag: continue # No suitable T-A pair was found on opposite sides of both fragments # Look for T-A pairs contained in either sequence # Distance between the T-A and proximity to the junction are important # factors T_positions_in_ths = [m.start() for m in re.finditer("T|t", str(ths.seq))] pairs_ths = product(A_positions_in_ths[::-1], T_positions_in_ths[::-1]) for ths_a, ths_t in pairs_ths: if ths_a > ths_t: continue sticky_length = ths_t - ths_a if sticky_length < min_overlap: continue if sticky_length > max_overlap: continue pair_ths = ths_a, ths_t break else: pair_ths = tuple() ths_a, ths_t = 0, 0 A_positions_in_nxt = [m.start() for m in re.finditer("A|a", str(nxt.seq))] pairs_nxt = product(A_positions_in_nxt, T_positions_in_nxt) for nxt_a, nxt_t in pairs_nxt: if nxt_a > nxt_t: continue sticky_length = nxt_t - nxt_a if sticky_length < min_overlap: continue if sticky_length > max_overlap: continue pair_nxt = nxt_a, nxt_t break else: pair_nxt = tuple() nxt_a, nxt_t = 0, 0 if (pair_ths and not pair_nxt) or len(ths) - ths_a <= nxt_t: # T-A pair in ths; # Move ths reverse primer downstream # Extend nxt foward primer tail fp = bytearray(ths.seq[ths_a:]._data + nxt.forward_primer.seq._data) fp[ths_t - ths_a] = ord(b"U") nxt.forward_primer = Primer(fp) shorter_ths = ths[: ths_t + 1] rp = bytearray( primer_design( shorter_ths, limit=ths_t - ths_a + 1 ).reverse_primer.seq._data ) rp[ths_t - ths_a] = ord(b"U") ths.reverse_primer = Primer(rp) elif (not pair_ths and pair_nxt) or len(ths) - ths_a >= nxt_t: # T-A pair in nxt; modify ths reverse primer # Move nxt forward primer upstream # Extend ths reverse primer tail rp = bytearray( nxt.seq[: nxt_t + 1].rc()._data + ths.reverse_primer.seq._data ) rp[nxt_t - nxt_a] = ord(b"U") ths.reverse_primer = Primer(rp) shorter_nxt = nxt[nxt_a:] fp = bytearray( primer_design( shorter_nxt, limit=nxt_t - nxt_a + 1 ).forward_primer.seq._data ) fp[nxt_t - nxt_a] = ord(b"U") nxt.forward_primer = Primer(fp) return amplicons