AACR new biologics sequences: a brief analysis primer

So, following on from the AACR 2026 disclosures of the small molecules we can now turn to the important topic of the sequences of the biologics. The first aspect to note is that GtoPdb generally only curates approved clinical biologics (predominantly antibodies) so the six disclosed in this session are not yet included. I have some experience in searching patent sequences (mostly targets) back in the day at both SmithKline and AstraZenea who both subscribed to Derwent GENSEQ (see also this older non-pub article on Patent Sequence Databases) so I know some of the quirks and challenges.  First up I have to apologise for not being able to retrospectively find an open link to the very useful PDF I stumbled upon wherein the sequences were listed (e.g. as per below) so I'd be pleased to add this if anyone has it.    

1)  AZD1221 (not to be confused with their covid vaccine  AZD1222)  is an AstraZeneca dual-payload antibody-drug conjugate (ADC) designed to target both microtubules (MT) and DNA topoisomerase I (TOP1).  We can find two sequences, presumably the light and heavy chains, below  

DIVLTQSPASLAVSLGQRATISCKASQSVDFDGDSYMNWYQQKPGQP

PKVLIYAASNLESGIPARFSGSGSGTDFTLNIHPVEEEDAATYYCQQSN

EDPWTFGGGTKLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYP

REAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLSSTLTLSKADYEK

HKVYACEVTHQGLSSPVTKSFNRGEC

QIQLQQSGPEVVKPGASVKISCKASGYTFTDYYITWVKQKPGQGLEWI

GWIYPGSGNTKYNEKFKGKATLTVDTSSSTAFMQLSSLTSEDTAVYFC

ANYGNYWFAYWGQGTQVTVSAASTKGPSVFPLAPSSKSTSGGTAAL

GCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPS

SSLGTQTYICNVNHKPSNTKVDKKVEPKSCDKTHTCPPCPAPELLGGP

SVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVH

NAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIE

KTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWE

SNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVM

HEALHNHYTQKSLSLSPGK

For using open sources a useful first-stop is to check NCBI non-redundant via BLASTP  for non-patent hits, starting with the longest chain

We can immediately perceive, at ~95% ID 1) there are no exact matches, 2) yes its a heavy chain, and thus 3) the mouse orthologue is likely to be "humanised" as a patent sequence.  The next step is to pop against NCBI "pataa" Protein sequences derived from the Patent division of GenBank, Update:2026/07/02, Number of sequences:3958113, with this result on 6th July;