RidZdddddZdddd dd Zd d d ZddddddddZdZGddZedk(rddlm Z e yy)aRCalculate isoelectric points of polypeptides using methods of Bjellqvist. pK values and the methods are taken from:: * Bjellqvist, B.,Hughes, G.J., Pasquali, Ch., Paquet, N., Ravier, F., Sanchez, J.-Ch., Frutiger, S. & Hochstrasser, D.F. The focusing positions of polypeptides in immobilized pH gradients can be predicted from their amino acid sequences. Electrophoresis 1993, 14, 1023-1031. * Bjellqvist, B., Basse, B., Olsen, E. and Celis, J.E. Reference points for comparisons of two-dimensional maps of proteins from different human cell types defined in a pH scale where isoelectric points correlate with polypeptide compositions. Electrophoresis 1994, 15, 529-539. I designed the algorithm according to a note by David L. Tabb, available at: http://fields.scripps.edu/DTASelect/20010710-pI-Algorithm.pdf g@g$@g(@gQ@)NtermKRHgffffff @333333@g@g"@)CtermDECYg333333@g@)rr g\(\@g@gQ@gQ @gHzG@g(\@g@)AMSPTVr )rrrrr r r c2eZdZdZddZdZdZdZd dZy) IsoelectricPointaA class for calculating the IEP or charge at given pH of a protein. Parameters ---------- :protein_sequence: A ``Bio.Seq`` or string object containing a protein sequence. :aa_content: A dictionary with amino acid letters as keys and its occurrences as integers, e.g. ``{"A": 3, "C": 0, ...}``. Default: ``None``. If ``None``, the dic will be calculated from the given sequence. Methods ------- :charge_at_pH(pH): Calculates the charge of the protein for a given pH :pi(): Calculates the isoelectric point Examples -------- The methods of this class can either be accessed from the class itself or from a ``ProtParam.ProteinAnalysis`` object (with partially different names): >>> from Bio.SeqUtils.IsoelectricPoint import IsoelectricPoint as IP >>> protein = IP("INGAR") >>> print(f"IEP of peptide {protein.sequence} is {protein.pi():.2f}") IEP of peptide INGAR is 9.75 >>> print(f"Its charge at pH 7 is {protein.charge_at_pH(7.0):.2f}") Its charge at pH 7 is 0.76 >>> from Bio.SeqUtils.ProtParam import ProteinAnalysis as PA >>> protein = PA("PETER") >>> print(f"IEP of {protein.sequence}: {protein.isoelectric_point():.2f}") IEP of PETER: 4.53 >>> print(f"Charge at pH 4.53: {protein.charge_at_pH(4.53):.2f}") Charge at pH 4.53: 0.00 Nc|j|_|s&ddlm}||jj }|j ||_|j\|_|_ y)zInitialize the class.)ProteinAnalysisN) uppersequenceBio.SeqUtils.ProtParamrcount_amino_acids_select_chargedcharged_aas_content_update_pKs_tablespos_pKsneg_pKs)selfprotein_sequence aa_content_PAs U/home/agent/.friday_env/lib/python3.12/site-packages/Bio/SeqUtils/IsoelectricPoint.py__init__zIsoelectricPoint.__init__RsY(..0  ET]]+==?J#'#7#7 #C %)%<%<%>" dlcVi}tD]}t||||<d|d<d|d<|S)N?rr) charged_aasfloat)r r"chargedaas r$rz IsoelectricPoint._select_charged_sA 0B 2/GBK 0r&ctj}tj}|jd|jd}}|tvr t||d<|t vr t ||d<||fS)z@Update pKs tables with seq specific values for N- and C-termini.rrr) positive_pKscopy negative_pKsr pKnterminal pKcterminal)r rrntermcterms r$rz#IsoelectricPoint._update_pKs_tablesgsr##%##%}}Q'r):u K *51GG  K *51GG r&c(d}|jjD](\}}dd||z zdzz }||j||zz }*d}|jjD](\}}dd||z zdzz }||j||zz }*||z S)z.Calculate the charge of a protein at given pH.r( )ritemsrr)r pHpositive_charger,pKpartial_chargenegative_charges r$ charge_at_pHzIsoelectricPoint.charge_at_pHrsll((* MFB B27Oc$9:N t77;nL LO Mll((* MFB B27Oc$9:N t77;nL LO M00r&c|j|}||z dkDr%|dkDr|}n|}||zdz }|j|||S|S)aCalculate and return the isoelectric point as float. This is a recursive function that uses bisection method. Wiki on bisection: https://en.wikipedia.org/wiki/Bisection_method Arguments: - pH: the pH at which the current charge of the protein is computed. This pH lies at the centre of the interval (mean of `min_` and `max_`). - min\_: the minimum of the interval. Initial value defaults to 4.05, which is below the theoretical minimum, when the protein is composed exclusively of aspartate. - max\_: the maximum of the the interval. Initial value defaults to 12, which is above the theoretical maximum, when the protein is composed exclusively of arginine. g-C6?r7)r?pi)r r:min_max_chargenext_pHs r$rBzIsoelectricPoint.pisW ""2& $; |d{a'G777D$/ / r&)N)g@r ) __name__ __module__ __qualname____doc__r%rrr?rBr&r$rr)s!&P ?  1.r&r__main__r) run_doctestN) rKr/r1r3r2r)rrH Bio._utilsrNrLr&r$rPs{&4d> Dt#DI t$  2 xxv z&Mr&