Open marcopretti opened 3 years ago
Note to myself: probably related to #131
Hello @Marco0107 and thank you so much for signing up on GitHub to open an issue, I truly appreciate it!
I'm deeply sorry for the very late response. We've been extremely busy raising investments and hiring the team after finishing UC Berkeley startup accelerator. We are finally set up for now: we grew the team from 4 to 15 amazing professionals, and raised a large investment round from great investors. Thank you so much for your patience!
We are thinking about CDR3-based clustering and fuzzy search, but we need more use cases and applications for it. Would you like to search using amino acid only, or do you consider something like BLOSUM-based alignment?
Improving upon Immunarch, we will soon release our online analytics platform for fast, effortless and code-less reproduction of impactul single-cell transcriptomics and immune repertoire publications. It's designed for researchers who spend a lot of time or energy reproducing sophisticated pipelines or analysing datasets from publications. Feel free to send us a link to publications of your interest so we can reproduce them on our platform and give you an access.
We are more than happy to grant you an early access as our gratitude for you being a longstanding immunarch user.
We have a list of datasets from around 30 publications that we are going to make available for your exploration and analysis. As an example, we recently reproduced on a platform a paper focused on scRNAseq-based estimation of efficacy of CAR-T immunotherapy.
Please send links to support@immunomind.io. The direct link to the publication datasets would be also greatly appreciated.
Don't worry @vadimnazarov and thank you for your response.
I consider search for amino acid only since really similar sequences could be recognizing the same epitope. BLOSUM-based alignment wouldn't be of interest (mine at least). I believe more than 3-4 divergent amino acids from the query sequence would be too much (to be discutted, depends on the position). However, CDR3-based clustering sounds really interesting..
I'm happy to see you are improving the tool and I can't wait to see it ;)
🚀 Feature
Hi guys, I'm new to github, so not sure how/where to post things. I believe a function that uses Levenshtein distance or others would be really useful for the package
Motivation
Some online databases offer the search for CDR3 sequences allowing one or more substitutions in the sequence. I like the usability of the dbAnnotate function, but I needed one that allowed substitutions/mismatches.
Alternatives
For that reason, I wrote this small piece of code below. I'm not an expert in R and sure the code can be improved. I'm sharing so maybe you could implement something similar.