Fig 1.
An overview of the data analysis pipeline used.
Fig 2.
Illustrations for hypotheses of different immune escape patterns.
The illustrations shown are for: an immune escape (A), reversion of an immune escape (B), and reversion of a zoonotic escape mutation (C). Grey curve = sequences with wuhan-1 peptide and dotted curve = sequences with newly emerged peptide with variant residue at the HLA anchor position.
Fig 3.
Escape patterns associated with HLA- A*66:01/A*68:01.
The proportions are shown as follows in all graphs: Wuhan residue = solid grey area; predicted escape variant = dotted area; predicted compensatory mutations = different styles of line graphs; escaping anchor position = bold-face font residue in the peptide; compensatory position = underlined residue in the peptide. A) Toggling between predicted escape and the Wuhan residue at second anchor position of the HLA- A*68:01 binding peptide ’94-STEKSNIIR-102’. The escape variant 95I also binds to the same HLA allele. DRB1*04:01 is associated with same T95I escape in 90-VYFASTEKS-98 (15mer = 87-NDGVYFASTEKSNII). B) HLA-A*66:01 and A*68:01 drive escape at 408, compensation 405N alone fails, but compensation with escape increase. DQA1*01:02-DQB1*06:02 binds the emerged 408S escape with 405N compensation. C) Proportions of predicted toggling escape at second anchor position of 370-NSASFSTFK-378 peptide binder to HLA- A*03:01 and A*66:01, with compensatory mutation sites 373 and 375. The escape mutation S371F dominates with additional compensation of T376A. D) Toggling between predicted escape and the Wuhan residue at C-term anchor position for the HLA- A*03:01 and A*66:01 binding peptide ’409-QIAPGQTGK-417’. The escape variant 417N is not a binder. E) N440K predicted reversion of zoonotic immune escape. The variant 440K is a binding C-term anchor motif for HLA- A*03:01/ A*66:01/ A*68:01. F) HLA-A*02:06/ A*66:01/ A*68:01 bind to the 769V variant. Wuhan likely zoonotic escape, with sporadic reversion attempts. DQA1*01:02-DQB1*06:02 also binds to 769V variant but DRB1*04:01 binds to Wuhan in overlapping 764-NRALTGIAV-772 (15mer = 761-TQLNRALTGIAVEQD). N-terminal N764K anchor mutation also observed and both binders to HLA-II. G) N856K attempted reversion of zoonotic immune escape at a cryptic anchor motif for HLA-B*08:01. The variant peptide ‘852-AQKFKGLTV-860’ is a strong binder but the Wuhan peptide is not a binder. The variant is also a C-term anchor for an overlapping A*66:01 binding peptide 848-DLICAQKFK-856.
Fig 4.
Escape patterns associated with other different HLA alleles.
The proportions are shown as follows in all graphs: Wuhan residue = solid grey area; predicted escape variant = dotted area; predicted compensatory mutations = different styles of line graphs; insertions = dotted bar graphs; escaping anchor position = bold-face font residue on the peptide; compensatory position = underlined residue on the peptide. A) HLA-B*08:01 binds Wuhan and both P681R/H escape variants including the later N679K compensation. B*56:01 binds the overlapping Wuhan peptide and a short-lived A688V C-term escape. Peptide overlaps with the C-term of DRB5*01:01 motif and 681 is adjacent to the C-term anchor in 674-YQTQTNSPR-682 (15mer = 671-CASYQTQTNSPRRAR), variants Q675H & H677Q, T678I are also binders. B) Toggling of attempted escape variants driven by B*27:02 and B*27:03 which both bind to the Wuhan peptide and 346T variant. B*27:02 also binds to the 346K variant. C) Proportions of predicted escape, reversion and compensation at C-terminal anchor position for HLA- A*02:05 binding peptide ‘10-LVSSQCVNL-18’. D) HLA-A*02:05 and several A*02 type alleles and/or numerous C alleles drive escape but also bind escape variant due to a supermotif. E) Toggling between predicted escape and the Wuhan residue at C-term anchor position for the HLA- A*02:05, A*02:06, A*02:14 and A*30:01 binding peptide’444-KVGGNYNYL-452’. The escape variant 452R is a binder to A*30:01, A*03:01 and A*68:01. F) Proportions of predicted escape V213G and fixation at second anchor of 212-LVRDLPQGF-220 peptide binder to HLA-B*15:02 and B*07:02, with toggling of compensatory mutation sites L212I/S and D215G within the peptide. A 215-EPE-217 insertion during wave-4 is a potential escape mechanism. DQA1*01:02-DQB1*06:02 binds 215G compensatory mutation in variant 207-HTPINLVRG-215 (15mer = 204-YSKHTPINLVRGLPQ). G) Overlapping Wuhan peptides. C-term F490S mutations short-lived and binds A*24:02/A*26:01. Compensation toggling between E484K/A + T478K. Wuhan and compensations are binders to B*07:02 or B*15:08 or B*35:01. Peptide 478-TPCNGVEGF-486 has C-term F486V escape from B*15:08/B*35:01 motifs.
Fig 5.
Zoonosis related escape patterns associated with other different HLA alleles.
The proportions are shown as follows in all graphs: Wuhan residue = solid grey area; predicted escape variant = dotted area; predicted compensatory mutations = different styles of line graphs; A) N501Y predicted reversion of zoonotic immune escape at C-term anchor position of a supermotif in the HLA- B*15:16/ B*15:17/ B*57:01/ B*58:01/A*03:01 binding variant peptide ‘493-QSYGFQPTY-501’. The mutation has toggling compensation while it also acts as a compensation for an overlapping peptide 495-YGFQPTNGV-503. DRB1*01:01 binds N501 and the variant 501Y, escape is at the compensatory positions 496S and 498R, 498 is an anchor in 495-YGFQPTNGV-503 (15mer = 492-LQSYGFQPTNGVGYQ). B) L981F attempted escape mutation or reversion of zoonotic escape mutation at C-term of 973-ISSVLNDIL-981 peptide. The variant is a strong binder to B*58:01/B*57:02. The Wuhan peptide was predicted as a weak binder to same alleles but in absence of known anchor residue motif. DRB1*11:04 also binds same Wuhan and variant on the N-term anchor in 981-LSRLDKVEA-989 (15mer = 978-NDILSRLDKVEAEVQ). C) A701V predicted attempted reversion of a zoonotic immune escape at C-term anchor position of 693-IAYTMSLGV-701 variant peptide with a strong binding motif for HLA- B*51:01/ B*51:02/ B*51:03 alleles. The Wuhan peptide is not a binder to any HLA. DRB1*07:01 binds both Wuhan (weak) and variant (strong) in C-term anchor of 693-IAYTMSLGA -701 (15mer = 690-QSIIAYTMSLGAENS).