-
Notifications
You must be signed in to change notification settings - Fork 43
/
Copy pathtodo.txt
218 lines (179 loc) · 9.07 KB
/
todo.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
- add eff score from https://github.com/zhangchonglab/sgRNA-cleavage-activity-prediction
was made for bacteria.
Crispor Edit:
- (C>A)
- (CCA>TGA)
- (CCA>del)
- (ins>CCA)
Indel predictions:
- Kim 2020: https://www.nature.com/articles/s41551-019-0505-1#code-availability https://github.com/CRISPRJWCHOI/IndelSearcher
- Wang 2019 https://www.nature.com/articles/s41467-019-12281-8 www.deephf.com
- Chakrabarti et al Mol Cell 2018 https://www.sciencedirect.com/science/article/pii/S1097276518310013?via%3Dihub
- Wei of Jay Shendure's group
Add more deletion predictors:
- https://www.nature.com/articles/nbt.4317
- https://partslab.sanger.ac.uk/FORECasT
Make the new TATV Cpf1 PAM work, cite 10.1038/nbt.3900 when done.
- Ajouter explications pourquoi saCas9. -> manuel -> warning
- Ajouter warning pour Cpf1, ajouter TTTV
JP:
IL ya aussi un outil tout bête à faire : quand on veut faire un KI, il faut faire une coupure le plus près possible du site d'insertion.
Donc ça serait utile d'aider à voir tout ce qui est possible comme coupures.
Ce qu'on fait aujourdh'ui, c'est d'utiliser S pyogenes/NGG et si ça ne va pas on cherche une autre protéine Cas avec un PAM mieu placé.
POur éviter d'avoir à tester chaque PAM dans la liste, on pourrait afficher (pour les toutes petites séquences input) tout les PAM possibles ! Par exemple ça serait surement suffisant de se limiter à moins de 70 pb. Qu'est ce que tu en penses ?
firefox: make sure that people can select text + indicate scroling is possible
baseeditor window has to be user-configurable, textbox, "x-y"
- baseeditor mode does not need to show PAMs
- BE mode: restriction enzymes are different
color off-targets by whether they are in a COSMIC gene or not.
Color all scores and add Jensen et al :
- p.s. red/yellow/green indicators would also help for predicted effeciency / out of frame scores.
- > One thing that I look for, which is not shown here is a test for self-complimentarity.
- I've often hear this and probably should add it. Is this the mfold score as published in Jenssen et al? Do you want to see the RNAfold diagram, too?
- Mfold score would be fine; if you could code it red/yellow/green to indicate good/bad scores that would help too.
- cfp1 support
- show all guides for all PAMs together
- suppress all eff scores but Fusi and Moreno
- knock-in support
Add these new criteria:
- Thyme/Schier: http://chopchop.cbu.uib.no/scoring.php, add a column?
see https://www.ncbi.nlm.nih.gov/pubmed/27282953
- Jensen et al: Chromatin and secondary structure
http://onlinelibrary.wiley.com/doi/10.1002/1873-3468.12707/full
- Jennifer's zebrafish model
- CRISPROff and CRISPRSpec? http://rth.dk/resources/crispr/
Show a histogram of the mismatches instead of 0-0-10-15-40
What is site-seq and circle-seq?
- add note to crispor batch and sat mut assistant about meaning of barcode?
analyzers:
crispresso
https://bioconductor.org/packages/release/bioc/html/CrispRVariants.html
http://www.rgenome.net/cas-analyzer/#!
http://www.nature.com/articles/srep30330
https://www.ncbi.nlm.nih.gov/pubmed/28333914 file S3
Cas-Analyzer http://bioinformatics.oxfordjournals.org/content/early/2016/09/15/bioinformatics.btw561.abstract
https://github.com/TuanjieNew/Indel-Finder
https://github.com/mlafave/ampliconDIVider
screen analyzers:
https://github.com/boutroslab/CRISPRAnalyzeR
screens CrisprI/A
https://github.com/mhorlbeck/CRISPRiaDesign
databases of tested guides:
- addgene https://www.addgene.org/crispr/reference/grna-sequence/
----
off-target searchers:
- https://github.com/htgt/CRISPR-Analyser
- breaking Cas
- https://github.com/MicrosoftResearch/Azimuth
- align against ALL human genomes: https://github.com/emptyewer/MAPster
- Church lab (CRISPR-GA): http://54.80.152.219/index.php
- on-target scores:
- https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-017-1697-6
Says that dinucleotides improve the AUC, code at:
http://www.ams.sunysb.edu/~pfkuan/softwares.html#predictsgrna
Good R code, easy to run
- CrisprPred http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0181943
code at: https://github.com/khaled-rahman/CRISPRpred
Weird R code, nothing easy to calculate the score, looks like CS people
No updates ever
- CCTop crisprRater https://academic.oup.com/nar/article/46/3/1375/4754467
off-targets:
- OfftargetPredict https://academic.oup.com/bioinformatics/article/34/17/i757/5093213
code at https://github.com/penn-hui/OfftargetPredict
- crisproff, unpublished but on github
induced deletions:
- see JP email, CSHL 2018
---
auto-downloader:
- download genome to "genome"
- download gff to "gff"
- check if filetype is gz, extract
- check if type is really fasta
- check if annot is really gff
- check if sequences do not contain spaces
- try to pick out a certain field from the fasta ID lines, is possible
- translate genome to 2bit
- create chrom.sizes
- check if gff-chroms match genome chromes
- if not: skip gff
- do the gff translation, if error: skip gff
- index the fasta
- keep a log of everything
- on success, reindex genomes and send out the log to email
- use rabbitmq
- split on " ", or ";"
- feature UCSC more prominently:
- output page must activate the track in the link to UCSC!
- show link maybe even before the job is submitted
- "enter gene name" -> redirect to UCSC
- show all guides for all PAMs together
- multi-PAM: "need alternative PAM" ? show them in grey
- score CrisprA/CrisprI
- design a library for a list of genes
- library design: input: list of genes, how many guides per gene, annotate with libraries?
- output as excel: geneSym, guideSeq, transcripts, % in CDS, scores, guideSeq
- also on guide overview page: show which ones are in gecko/brunello etc
- more precalc of guides / whole genome -> library design
- knock-in support -> concentrate onto GFP-fusions
- knock-in support
- what are the typical use cases: promoter/enhancer/knock-in/knock-out (=exon-removal)
- add-genome tool
- base de donnes:
-- text mining?
-- guides valides des cribles: ajouter si le guide etait deja utilise
(Sanger? Base de NIH zebrafish?)
fournir dans les tracks UCSC les guides des gènes essentiels déjà testés dans les cribles et renvoyer un lien
à GenomeCRISPR de Heidelberg pour plus détails ?
Analyses:
- variantes Cas9: kleinstiver VQR guideSeq <-> CFD ?
me semble vraiment important et original !
c’est par exemple possible de comparer et confirmer la supériorité de CFD
sur les MIT score pour ces variants.
- Cpf1: guideSeq <-> CFD ?
pour Cpf1 il fait designer un score à la main, comme CCTOP par exemple .
- screens:
- TKO http://tko.ccbr.utoronto.ca/
- https://elifesciences.org/articles/12677/figures
- Croatan guides:
http://www.cell.com/molecular-cell/abstract/S1097-2765(17)30464-1
already in ~/croatanGuides.txt
protocol in http://croatan.hannonlab.org/files/Supplementary_Protocol.pdf
- how many published guides may be affected by variants in off-targets?
pas évident, ça doit être facile d’identifier tous les variants qui diminuent
l’activité de off^targets (en passant de n mismatches à n+1) dans le sens
inverse, c’est plus compliqué de trouver les variants qui augmentent
l’activité de offtargets (en passant de n mismatches à n-1) puisqu’il faut
faire les calculs pour avoir tous les off avec 5mm et voir ceux dont les
variant font passer à 4 mm.
Done:
- cfp1 support
- suppress all eff scores but Fusi and Moreno
- remove Tefor menu from Pcr pages
- use case: hammer a region (enhancer) with guides: saturating mutagenesis libraries
- serial cloner support
- activate repeat filtering for the primer design !
(find the option in primer3 to remove low-complexity sequences)
To: Zachary Gilbert
Awesome! That was really fast! I just tested it out with a region we used
before and it gave us the same guides that worked well for us when we tested
them in mosquitoes. When I started looking for guides, about a year ago, I used
Chop-Chop and Crispr.MIT.edu. More recently, I have been using Benchling
because the interface was a little simpler, but none of them are updating
genomes right now, and the Aedes genome 5 is much more representative of the
actual genome than version 3, which is important to us, since we have had a
few occasions where we were trying guides in regions with a lot of SNPs, so not
many of them worked well. We also tried a few downloadable tools, although I
never got very far with them, since I ran into some weird permissions and
compatibility problems on our server.
For my needs, I prefer being able to enter a sequence and have the tool give me
guides in that sequence to choose from, the way it does now, since we often
want to, for example, find a guide very close to the stop codon, so we can
label a gene while, mostly, maintaining its function. Some of my lab mates,
though, would want to be able to enter a gene name and have it spit out a list
of good guides in the first and second exon of a gene, for creating knockouts.
Thanks again, this will be super helpful for finding guides for our next few
projects
Zach
Done:
Show distance of off-targets from the on-target!