pMHC movement based on peptide anchoring

Introduction

In this notebook, apo-holo comparisons between different anchoring patterns of MHC alleles plotted next to each other. The anchoring patterns were determined by finding the peptide motifs of each MHC allele from the MHCMotifAtlas and using these strong motifs as anchors for the peptide.

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import seaborn as sns
from python_pdb.formats.residue import THREE_TO_ONE_CODE

mhc_anchor_position_df = pd.read_csv('../data/external/mhc_motif_atlas.csv')
mhc_anchor_position_df = mhc_anchor_position_df.query("grade in ('high', 'dominant')")
mhc_anchor_position_df = mhc_anchor_position_df.sort_values('allele_slug')
mhc_anchor_position_df = mhc_anchor_position_df[['allele_slug', 'position', 'amino_acid', 'peptide_length']]
mhc_anchor_position_df = mhc_anchor_position_df.reset_index(drop=True)

mhc_anchor_position_df

	allele_slug	position	amino_acid	peptide_length
0	hla_a_01_01	2	T	9
1	hla_a_01_01	3	D	9
2	hla_a_01_01	9	Y	9
3	hla_a_02_01	2	L	9
4	hla_a_02_01	9	L	9
...	...	...	...	...
319	hla_g_01_03	1	K	9
320	hla_g_01_04	1	R	9
321	hla_g_01_04	3	P	9
322	hla_g_01_04	1	K	9
323	hla_g_01_04	9	L	9

324 rows × 4 columns

Load apo-holo comparisons

apo_holo_comparison = pd.read_csv('../data/processed/apo-holo-tcr-pmhc-class-I-comparisons/pmhc_per_res_apo_holo.csv')

peptide_apo_holo_comparison = apo_holo_comparison.query("chain_type == 'antigen_chain'").copy()
peptide_apo_holo_comparison

	complex_id	structure_x_name	structure_y_name	chain_type	residue_name	residue_seq_id	residue_insert_code	rmsd	ca_distance	chi_angle_change	com_distance
181	5c0a_D-E-C-A-B_tcr_pmhc	5c0a_D-E-C-A-B_tcr_pmhc.pdb	5n1y_C-A-B_pmhc.pdb	antigen_chain	MET	1	NaN	0.700531	0.237162	1.570636	0.364628
182	5c0a_D-E-C-A-B_tcr_pmhc	5c0a_D-E-C-A-B_tcr_pmhc.pdb	5n1y_C-A-B_pmhc.pdb	antigen_chain	VAL	2	NaN	0.114569	0.089359	0.009407	0.043732
183	5c0a_D-E-C-A-B_tcr_pmhc	5c0a_D-E-C-A-B_tcr_pmhc.pdb	5n1y_C-A-B_pmhc.pdb	antigen_chain	TRP	3	NaN	0.400840	0.233532	-0.019115	0.363989
184	5c0a_D-E-C-A-B_tcr_pmhc	5c0a_D-E-C-A-B_tcr_pmhc.pdb	5n1y_C-A-B_pmhc.pdb	antigen_chain	GLY	4	NaN	0.495618	0.262271	NaN	0.448252
185	5c0a_D-E-C-A-B_tcr_pmhc	5c0a_D-E-C-A-B_tcr_pmhc.pdb	5n1y_C-A-B_pmhc.pdb	antigen_chain	PRO	5	NaN	0.734430	0.486409	-0.842595	0.536198
...	...	...	...	...	...	...	...	...	...	...	...
209372	7rtr_D-E-C-A-B_tcr_pmhc	7rtd_C-A-B_pmhc.pdb	7rtr_D-E-C-A-B_tcr_pmhc.pdb	antigen_chain	ARG	5	NaN	1.064663	0.157924	0.364432	0.705033
209373	7rtr_D-E-C-A-B_tcr_pmhc	7rtd_C-A-B_pmhc.pdb	7rtr_D-E-C-A-B_tcr_pmhc.pdb	antigen_chain	THR	6	NaN	0.421897	0.345439	0.097701	0.344074
209374	7rtr_D-E-C-A-B_tcr_pmhc	7rtd_C-A-B_pmhc.pdb	7rtr_D-E-C-A-B_tcr_pmhc.pdb	antigen_chain	PHE	7	NaN	1.225982	0.317819	-0.347835	0.883236
209375	7rtr_D-E-C-A-B_tcr_pmhc	7rtd_C-A-B_pmhc.pdb	7rtr_D-E-C-A-B_tcr_pmhc.pdb	antigen_chain	LEU	8	NaN	1.323615	0.310356	0.612672	0.270049
209376	7rtr_D-E-C-A-B_tcr_pmhc	7rtd_C-A-B_pmhc.pdb	7rtr_D-E-C-A-B_tcr_pmhc.pdb	antigen_chain	LEU	9	NaN	1.599917	0.192969	0.885323	0.175807

9857 rows × 11 columns

peptide_apo_holo_comparison['peptide_length'] = \
    peptide_apo_holo_comparison.groupby(['complex_id',
                                         'structure_x_name',
                                         'structure_y_name']).transform('size')

peptide_apo_holo_comparison['amino_acid'] = peptide_apo_holo_comparison['residue_name'].map(THREE_TO_ONE_CODE)

Load summary data

summary_df = pd.read_csv('../data/processed/apo-holo-tcr-pmhc-class-I/apo_holo_summary.csv')
summary_df['group_name'] = summary_df['file_name'].str.replace('.pdb', '')

summary_df

/var/scratch/bmcmaste/2178229/ipykernel_2204500/1956598565.py:2: FutureWarning: The default value of regex will change from True to False in a future version.
  summary_df['group_name'] = summary_df['file_name'].str.replace('.pdb', '')

	file_name	pdb_id	structure_type	state	alpha_chain	beta_chain	antigen_chain	mhc_chain1	mhc_chain2	cdr_sequences_collated	peptide_sequence	mhc_slug	group_name
0	1ao7_D-E-C-A-B_tcr_pmhc.pdb	1ao7	tcr_pmhc	holo	D	E	C	A	B	DRGSQS-IYSNGD-AVTTDSWGKLQ-MNHEY-SVGAGI-ASRPGLA...	LLFGYPVYV	hla_a_02_01	1ao7_D-E-C-A-B_tcr_pmhc
1	1b0g_C-A-B_pmhc.pdb	1b0g	pmhc	apo	NaN	NaN	C	A	B	NaN	ALWGFFPVL	hla_a_02_01	1b0g_C-A-B_pmhc
2	1b0g_F-D-E_pmhc.pdb	1b0g	pmhc	apo	NaN	NaN	F	D	E	NaN	ALWGFFPVL	hla_a_02_01	1b0g_F-D-E_pmhc
3	1bd2_D-E-C-A-B_tcr_pmhc.pdb	1bd2	tcr_pmhc	holo	D	E	C	A	B	NSMFDY-ISSIKDK-AAMEGAQKLV-MNHEY-SVGAGI-ASSYPGG...	LLFGYPVYV	hla_a_02_01	1bd2_D-E-C-A-B_tcr_pmhc
4	1bii_P-A-B_pmhc.pdb	1bii	pmhc	apo	NaN	NaN	P	A	B	NaN	RGPGRAFVTI	h2_dd	1bii_P-A-B_pmhc
...	...	...	...	...	...	...	...	...	...	...	...	...	...
386	7rtd_C-A-B_pmhc.pdb	7rtd	pmhc	apo	NaN	NaN	C	A	B	NaN	YLQPRTFLL	hla_a_02_01	7rtd_C-A-B_pmhc
387	7rtr_D-E-C-A-B_tcr_pmhc.pdb	7rtr	tcr_pmhc	holo	D	E	C	A	B	DRGSQS-IYSNGD-AVNRDDKII-SEHNR-FQNEAQ-ASSPDIEQY	YLQPRTFLL	hla_a_02_01	7rtr_D-E-C-A-B_tcr_pmhc
388	8gvb_A-B-P-H-L_tcr_pmhc.pdb	8gvb	tcr_pmhc	holo	A	B	P	H	L	YGATPY-YFSGDTLV-AVGFTGGGNKLT-SEHNR-FQNEAQ-ASSD...	RYPLTFGW	hla_a_24_02	8gvb_A-B-P-H-L_tcr_pmhc
389	8gvg_A-B-P-H-L_tcr_pmhc.pdb	8gvg	tcr_pmhc	holo	A	B	P	H	L	YGATPY-YFSGDTLV-AVGFTGGGNKLT-SEHNR-FQNEAQ-ASSD...	RFPLTFGW	hla_a_24_02	8gvg_A-B-P-H-L_tcr_pmhc
390	8gvi_A-B-P-H-L_tcr_pmhc.pdb	8gvi	tcr_pmhc	holo	A	B	P	H	L	YGATPY-YFSGDTLV-AVVFTGGGNKLT-SEHNR-FQNEAQ-ASSL...	RYPLTFGW	hla_a_24_02	8gvi_A-B-P-H-L_tcr_pmhc

391 rows × 13 columns

Annotate apo-holo data with allele information

peptide_apo_holo_comparison = peptide_apo_holo_comparison.merge(summary_df, how='left', left_on='complex_id', right_on='group_name')
peptide_apo_holo_comparison

	complex_id	structure_x_name	structure_y_name	chain_type	residue_name	residue_seq_id	residue_insert_code	rmsd	ca_distance	chi_angle_change	...	state	alpha_chain	beta_chain	antigen_chain	mhc_chain1	mhc_chain2	cdr_sequences_collated	peptide_sequence	mhc_slug	group_name
0	5c0a_D-E-C-A-B_tcr_pmhc	5c0a_D-E-C-A-B_tcr_pmhc.pdb	5n1y_C-A-B_pmhc.pdb	antigen_chain	MET	1	NaN	0.700531	0.237162	1.570636	...	holo	D	E	C	A	B	NSAFQY-TYSSGN-AMRGDSSYKLI-SGHDY-FNNNVP-ASSLWEK...	MVWGPDPLYV	hla_a_02_01	5c0a_D-E-C-A-B_tcr_pmhc
1	5c0a_D-E-C-A-B_tcr_pmhc	5c0a_D-E-C-A-B_tcr_pmhc.pdb	5n1y_C-A-B_pmhc.pdb	antigen_chain	VAL	2	NaN	0.114569	0.089359	0.009407	...	holo	D	E	C	A	B	NSAFQY-TYSSGN-AMRGDSSYKLI-SGHDY-FNNNVP-ASSLWEK...	MVWGPDPLYV	hla_a_02_01	5c0a_D-E-C-A-B_tcr_pmhc
2	5c0a_D-E-C-A-B_tcr_pmhc	5c0a_D-E-C-A-B_tcr_pmhc.pdb	5n1y_C-A-B_pmhc.pdb	antigen_chain	TRP	3	NaN	0.400840	0.233532	-0.019115	...	holo	D	E	C	A	B	NSAFQY-TYSSGN-AMRGDSSYKLI-SGHDY-FNNNVP-ASSLWEK...	MVWGPDPLYV	hla_a_02_01	5c0a_D-E-C-A-B_tcr_pmhc
3	5c0a_D-E-C-A-B_tcr_pmhc	5c0a_D-E-C-A-B_tcr_pmhc.pdb	5n1y_C-A-B_pmhc.pdb	antigen_chain	GLY	4	NaN	0.495618	0.262271	NaN	...	holo	D	E	C	A	B	NSAFQY-TYSSGN-AMRGDSSYKLI-SGHDY-FNNNVP-ASSLWEK...	MVWGPDPLYV	hla_a_02_01	5c0a_D-E-C-A-B_tcr_pmhc
4	5c0a_D-E-C-A-B_tcr_pmhc	5c0a_D-E-C-A-B_tcr_pmhc.pdb	5n1y_C-A-B_pmhc.pdb	antigen_chain	PRO	5	NaN	0.734430	0.486409	-0.842595	...	holo	D	E	C	A	B	NSAFQY-TYSSGN-AMRGDSSYKLI-SGHDY-FNNNVP-ASSLWEK...	MVWGPDPLYV	hla_a_02_01	5c0a_D-E-C-A-B_tcr_pmhc
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
9852	7rtr_D-E-C-A-B_tcr_pmhc	7rtd_C-A-B_pmhc.pdb	7rtr_D-E-C-A-B_tcr_pmhc.pdb	antigen_chain	ARG	5	NaN	1.064663	0.157924	0.364432	...	holo	D	E	C	A	B	DRGSQS-IYSNGD-AVNRDDKII-SEHNR-FQNEAQ-ASSPDIEQY	YLQPRTFLL	hla_a_02_01	7rtr_D-E-C-A-B_tcr_pmhc
9853	7rtr_D-E-C-A-B_tcr_pmhc	7rtd_C-A-B_pmhc.pdb	7rtr_D-E-C-A-B_tcr_pmhc.pdb	antigen_chain	THR	6	NaN	0.421897	0.345439	0.097701	...	holo	D	E	C	A	B	DRGSQS-IYSNGD-AVNRDDKII-SEHNR-FQNEAQ-ASSPDIEQY	YLQPRTFLL	hla_a_02_01	7rtr_D-E-C-A-B_tcr_pmhc
9854	7rtr_D-E-C-A-B_tcr_pmhc	7rtd_C-A-B_pmhc.pdb	7rtr_D-E-C-A-B_tcr_pmhc.pdb	antigen_chain	PHE	7	NaN	1.225982	0.317819	-0.347835	...	holo	D	E	C	A	B	DRGSQS-IYSNGD-AVNRDDKII-SEHNR-FQNEAQ-ASSPDIEQY	YLQPRTFLL	hla_a_02_01	7rtr_D-E-C-A-B_tcr_pmhc
9855	7rtr_D-E-C-A-B_tcr_pmhc	7rtd_C-A-B_pmhc.pdb	7rtr_D-E-C-A-B_tcr_pmhc.pdb	antigen_chain	LEU	8	NaN	1.323615	0.310356	0.612672	...	holo	D	E	C	A	B	DRGSQS-IYSNGD-AVNRDDKII-SEHNR-FQNEAQ-ASSPDIEQY	YLQPRTFLL	hla_a_02_01	7rtr_D-E-C-A-B_tcr_pmhc
9856	7rtr_D-E-C-A-B_tcr_pmhc	7rtd_C-A-B_pmhc.pdb	7rtr_D-E-C-A-B_tcr_pmhc.pdb	antigen_chain	LEU	9	NaN	1.599917	0.192969	0.885323	...	holo	D	E	C	A	B	DRGSQS-IYSNGD-AVNRDDKII-SEHNR-FQNEAQ-ASSPDIEQY	YLQPRTFLL	hla_a_02_01	7rtr_D-E-C-A-B_tcr_pmhc

9857 rows × 26 columns

Combine anchoring information with apo-holo comparisons

mhc_anchor_position_df['anchor'] = True

peptide_apo_holo_comparison = peptide_apo_holo_comparison.merge(
    mhc_anchor_position_df,
    how='left',
    left_on=['mhc_slug', 'residue_seq_id', 'amino_acid', 'peptide_length'],
    right_on=['allele_slug', 'position', 'amino_acid', 'peptide_length'],
)

def collate_anchors(group: pd.DataFrame) -> list[int]:
    return sorted(group[group['anchor'] == True]['residue_seq_id'].unique().tolist())

anchoring_strategies = peptide_apo_holo_comparison.groupby(['structure_x_name', 'structure_y_name']).apply(collate_anchors)
anchoring_strategies.name = 'anchoring_strategy'
anchoring_strategies = anchoring_strategies.reset_index()

peptide_apo_holo_comparison = peptide_apo_holo_comparison.merge(anchoring_strategies, how='left', on=['structure_x_name', 'structure_y_name'])

def markup_anchor(anchors: list[int]) -> str:
    anchors = [f'p{anchor}' for anchor in anchors]
    return '-'.join(anchors)

peptide_apo_holo_comparison['anchoring_strategy_str'] = peptide_apo_holo_comparison['anchoring_strategy'].map(markup_anchor)

def find_dominant_anchor(group: pd.DataFrame) -> str:
    anchor_types = group['anchoring_strategy_str'].unique()
    lengths = np.array([len(anchor) for anchor in anchor_types])
    index = np.argmax(lengths)

    return anchor_types[index]

dominant_anchors = peptide_apo_holo_comparison.groupby('mhc_slug').apply(find_dominant_anchor)
dominant_anchors.name = 'dominant_anchor'
dominant_anchors = dominant_anchors.reset_index()

peptide_apo_holo_comparison = peptide_apo_holo_comparison.merge(dominant_anchors, how='left', on='mhc_slug')

The below correction is applied as it is assumed that those peptide marked as solely ‘p2’ or ‘p9’ anchors are most likely still anchored by something at either end that is not necessarily the high or dominant motif.

peptide_apo_holo_comparison['dominant_anchor'] = peptide_apo_holo_comparison['dominant_anchor'].map(
    lambda strategy: 'p2-p9' if strategy in ('p2', 'p9') else strategy
)

peptide_apo_holo_comparison_with_anchor = peptide_apo_holo_comparison.query("anchoring_strategy_str != ''")

Visualising the results

g = sns.catplot(peptide_apo_holo_comparison_with_anchor,
                col='dominant_anchor',
                x='residue_seq_id', y='rmsd',
                sharex=False,
                color='salmon',
                kind='bar')

def annotate(data, **kws):
    groups = data.groupby(['mhc_slug', 'peptide_sequence'])
    ax = plt.gca()
    ax.text(0.45, 0.9, f'n={len(groups)}', fontsize=16, transform=ax.transAxes)

g.map_dataframe(annotate)

<seaborn.axisgrid.FacetGrid at 0x7f6443502380>

max_rmsd = peptide_apo_holo_comparison_with_anchor['rmsd'].max()

for mhc_slug, group in peptide_apo_holo_comparison_with_anchor.groupby('mhc_slug'):
    print(mhc_slug)
    print(group['dominant_anchor'].unique()[0])
    plot = sns.barplot(group, x='residue_seq_id', y='rmsd', color='salmon')
    plot.set_ylim((0, np.ceil(max_rmsd)))
    plt.show()

hla_a_02_01
p2-p9

hla_a_24_02
p2-p9

hla_b_07_02
p2-p9

hla_b_08_01
p2-p5-p9

hla_b_35_01
p2-p9

hla_b_37_01
p2-p5-p9

hla_b_42_01
p2-p9

hla_b_44_05
p2-p9

hla_b_53_01
p2-p9

hla_b_81_01
p2-p9

hla_e_01_03
p2-p9

sns.barplot(peptide_apo_holo_comparison_with_anchor,
            hue='dominant_anchor',
            x='residue_seq_id', y='rmsd')

<AxesSubplot: xlabel='residue_seq_id', ylabel='rmsd'>

representative_alleles = (peptide_apo_holo_comparison_with_anchor.groupby('dominant_anchor')[['dominant_anchor', 'mhc_slug']]
                                                                 .sample(1, random_state=1)
                                                                 .reset_index(drop=True))
representative_alleles

	dominant_anchor	mhc_slug
0	p2-p5-p9	hla_b_08_01
1	p2-p9	hla_e_01_03

sns.barplot(peptide_apo_holo_comparison_with_anchor.query("mhc_slug.isin(@representative_alleles['mhc_slug'])"),
            hue='mhc_slug',
            x='residue_seq_id', y='rmsd')

<AxesSubplot: xlabel='residue_seq_id', ylabel='rmsd'>

Conclusion

From the visualisations it is clear that the profile of peptide conformational changes depends on the mhc allele and how the peptides are anchored by the allele.