dsh-molbio-tools 0.19.0 → 0.20.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.md +100 -0
- package/README.md +65 -21
- package/align.mjs +28 -0
- package/batch.mjs +185 -0
- package/benchmark/README.md +27 -8
- package/benchmark/sequences.mjs +7 -0
- package/benchmark/tasks/02-primers.json +132 -19
- package/benchmark/tasks/03-restriction.json +113 -29
- package/benchmark/tasks/04-cloning.json +167 -34
- package/benchmark/tasks/05-plasmids.json +172 -29
- package/benchmark/tasks/06-analysis.json +387 -100
- package/benchmark/tasks/09-formats.json +132 -23
- package/benchmark/tools.mjs +28 -2
- package/benchmark/verifications.mjs +17 -1
- package/build/release-preflight.mjs +2 -2
- package/crispr.mjs +392 -78
- package/docs/history.md +10 -0
- package/docs/maintainer.md +8 -6
- package/docs/roadmap.md +22 -8
- package/docs/rules.md +1 -1
- package/docs/workflow.md +7 -7
- package/fastq-qc.mjs +735 -735
- package/index.mjs +573 -36
- package/lib/client.js +665 -86
- package/multiplex.mjs +87 -0
- package/package.json +9 -4
- package/packages/molbio-panel/lib/client.js +665 -86
- package/packages/molbio-panel/package.json +1 -1
- package/preset/molbio-lab/agent.cordis.yml +4 -4
- package/preset/molbio-lab/preset.patch.yml +2 -2
- package/preset/molbio-lab/preset.yml +1 -1
- package/silent-site.mjs +286 -0
- package/snapgene.mjs +17 -1
- package/taqman.mjs +163 -7
- package/test/contract.mjs +14 -0
- package/test/release.mjs +23 -3
- package/test/smoke.mjs +377 -3
- package/benchmark/reports/20260923-142935-suite.json +0 -545
- package/benchmark/reports/20260923-142935-suite.md +0 -30
- package/benchmark/reports/20260923-190149-suite.json +0 -2024
- package/benchmark/reports/20260923-190149-suite.md +0 -235
- package/benchmark/reports/20260923-192913-suite.json +0 -1856
- package/benchmark/reports/20260923-192913-suite.md +0 -158
- package/benchmark/reports/20260923-200212-suite.json +0 -3587
- package/benchmark/reports/20260923-200212-suite.md +0 -163
- package/benchmark/reports/20260923-202830-suite.json +0 -3129
- package/benchmark/reports/20260923-202830-suite.md +0 -99
- package/benchmark/reports/20260923-203856-suite.json +0 -71
- package/benchmark/reports/20260923-203856-suite.md +0 -16
|
@@ -13,95 +13,228 @@
|
|
|
13
13
|
"id": "region-extract",
|
|
14
14
|
"category": "cloning",
|
|
15
15
|
"tier": "full",
|
|
16
|
-
"covers": [
|
|
16
|
+
"covers": [
|
|
17
|
+
"molbio_extract_region"
|
|
18
|
+
],
|
|
17
19
|
"instruction": "Pull the region from position 101 to 200 out of this 512 bp sequence and give me the subsequence. Also confirm which restriction site sits at the very start of what you extracted.\n\n{seq:construct}",
|
|
18
20
|
"timeout_ms": 300000,
|
|
19
|
-
"expect_tools": [
|
|
21
|
+
"expect_tools": [
|
|
22
|
+
"molbio_extract_region"
|
|
23
|
+
],
|
|
20
24
|
"assertions": [
|
|
21
25
|
{
|
|
22
26
|
"where": "tool",
|
|
23
27
|
"pattern": "extracted (vector|sequence) 101-200 \\(100 bp\\)",
|
|
24
28
|
"note": "the tool ran with the sequence it was given. Measured behaviour worth knowing: in the full-tier run the model called `molbio_extract_region` THREE times and every call came back `Error: provide either `vector` (a sequence) or `vector_path`` — it never found the parameter, then reconstructed the region by hand and verified it with `molbio_align` (which it did correctly). This assertion is therefore a genuine TOOL-SELECTION/argument finding: the parameter is named `vector`, the description says 'a sequence', and the model reached for something else. Recorded in benchmark/README.md."
|
|
25
29
|
},
|
|
26
|
-
{
|
|
27
|
-
|
|
28
|
-
|
|
30
|
+
{
|
|
31
|
+
"where": "tool",
|
|
32
|
+
"pattern": "GAATTCAGACCTGCAATGGG",
|
|
33
|
+
"note": "the extracted subsequence starts with the EcoRI site"
|
|
34
|
+
},
|
|
35
|
+
{
|
|
36
|
+
"where": "answer",
|
|
37
|
+
"pattern": "EcoRI"
|
|
38
|
+
},
|
|
39
|
+
{
|
|
40
|
+
"where": "answer",
|
|
41
|
+
"pattern": "GAATTC"
|
|
42
|
+
}
|
|
29
43
|
]
|
|
30
44
|
},
|
|
31
45
|
{
|
|
32
46
|
"id": "mutagenesis",
|
|
33
47
|
"category": "cloning",
|
|
34
48
|
"tier": "full",
|
|
35
|
-
"covers": [
|
|
49
|
+
"covers": [
|
|
50
|
+
"molbio_mutagenesis_primers"
|
|
51
|
+
],
|
|
36
52
|
"instruction": "I want to change the A at position 200 of this template to a G by QuikChange-style site-directed mutagenesis. Design the primer pair and tell me what the change does to the protein.\n\n{seq:mutagenesis_template}",
|
|
37
53
|
"timeout_ms": 300000,
|
|
38
|
-
"expect_tools": [
|
|
54
|
+
"expect_tools": [
|
|
55
|
+
"molbio_mutagenesis_primers"
|
|
56
|
+
],
|
|
39
57
|
"assertions": [
|
|
40
|
-
{
|
|
41
|
-
|
|
42
|
-
|
|
43
|
-
|
|
44
|
-
{
|
|
58
|
+
{
|
|
59
|
+
"where": "tool",
|
|
60
|
+
"pattern": "mutation span 200-200 on the mutated template \\(400 bp\\)"
|
|
61
|
+
},
|
|
62
|
+
{
|
|
63
|
+
"where": "tool",
|
|
64
|
+
"pattern": "amino-acid change \\(frame 1 assumption\\): Q → R"
|
|
65
|
+
},
|
|
66
|
+
{
|
|
67
|
+
"where": "tool",
|
|
68
|
+
"pattern": "F CAGAGTGACAACGACATTGCTAGCC",
|
|
69
|
+
"note": "the designed forward primer"
|
|
70
|
+
},
|
|
71
|
+
{
|
|
72
|
+
"where": "tool",
|
|
73
|
+
"pattern": "R GGCTAGCAATGTCGTTGTCACTCTG"
|
|
74
|
+
},
|
|
75
|
+
{
|
|
76
|
+
"where": "answer",
|
|
77
|
+
"pattern": "Q.{0,6}R|Gln.{0,6}Arg",
|
|
78
|
+
"note": "the amino-acid consequence"
|
|
79
|
+
}
|
|
45
80
|
]
|
|
46
81
|
},
|
|
47
82
|
{
|
|
48
83
|
"id": "clone-primers",
|
|
49
84
|
"category": "cloning",
|
|
50
85
|
"tier": "full",
|
|
51
|
-
"covers": [
|
|
86
|
+
"covers": [
|
|
87
|
+
"molbio_clone_primers"
|
|
88
|
+
],
|
|
52
89
|
"instruction": "Design cloning primers to amplify this insert so I can ligate it into a vector cut with EcoRI and HindIII — add the sites plus protection bases. Tell me the full primer sequences, and warn me about anything that would go wrong in the digest.\n\n{seq:clone_insert}",
|
|
53
90
|
"timeout_ms": 300000,
|
|
54
|
-
"expect_tools": [
|
|
91
|
+
"expect_tools": [
|
|
92
|
+
"molbio_clone_primers"
|
|
93
|
+
],
|
|
55
94
|
"assertions": [
|
|
56
|
-
{
|
|
57
|
-
|
|
58
|
-
|
|
95
|
+
{
|
|
96
|
+
"where": "tool",
|
|
97
|
+
"pattern": "cloning primers \\(restriction\\)"
|
|
98
|
+
},
|
|
99
|
+
{
|
|
100
|
+
"where": "tool",
|
|
101
|
+
"pattern": "GATAGAATTC",
|
|
102
|
+
"note": "the forward primer's protection bases + EcoRI site"
|
|
103
|
+
},
|
|
104
|
+
{
|
|
105
|
+
"where": "tool",
|
|
106
|
+
"pattern": "AATAAAGCTT",
|
|
107
|
+
"note": "the reverse primer's tail"
|
|
108
|
+
},
|
|
59
109
|
{
|
|
60
110
|
"where": "tool",
|
|
61
111
|
"pattern": "WARNING: EcoRI also cuts INSIDE the template",
|
|
62
112
|
"note": "the insert carries an internal EcoRI site — the tool says so, and relaying it is the point of the task"
|
|
63
113
|
},
|
|
64
|
-
{
|
|
65
|
-
|
|
66
|
-
|
|
114
|
+
{
|
|
115
|
+
"where": "answer",
|
|
116
|
+
"pattern": "GAATTC"
|
|
117
|
+
},
|
|
118
|
+
{
|
|
119
|
+
"where": "answer",
|
|
120
|
+
"pattern": "AAGCTT"
|
|
121
|
+
},
|
|
122
|
+
{
|
|
123
|
+
"where": "answer",
|
|
124
|
+
"pattern": "inside|internal|would be digested|cuts"
|
|
125
|
+
}
|
|
67
126
|
]
|
|
68
127
|
},
|
|
69
128
|
{
|
|
70
129
|
"id": "clone-simulate",
|
|
71
130
|
"category": "cloning",
|
|
72
131
|
"tier": "full",
|
|
73
|
-
"covers": [
|
|
132
|
+
"covers": [
|
|
133
|
+
"molbio_clone_simulate"
|
|
134
|
+
],
|
|
74
135
|
"instruction": "Simulate a restriction cloning reaction: use EcoRI and HindIII to put this insert into my 512 bp vector. Add the enzyme sites to the insert for me. Tell me the size of the final plasmid and the fragments each enzyme gives on the finished construct.\n\nVector:\n{seq:construct}\n\nInsert:\n{seq:clone_insert}",
|
|
75
136
|
"timeout_ms": 300000,
|
|
76
|
-
"expect_tools": [
|
|
137
|
+
"expect_tools": [
|
|
138
|
+
"molbio_clone_simulate"
|
|
139
|
+
],
|
|
77
140
|
"assertions": [
|
|
78
|
-
{
|
|
79
|
-
|
|
80
|
-
|
|
81
|
-
|
|
82
|
-
{
|
|
83
|
-
|
|
141
|
+
{
|
|
142
|
+
"where": "tool",
|
|
143
|
+
"pattern": "restriction clone simulated: final plasmid 624 bp \\(Δ \\+112 bp\\)"
|
|
144
|
+
},
|
|
145
|
+
{
|
|
146
|
+
"where": "tool",
|
|
147
|
+
"pattern": "enzyme recognition sites were added to the bare insert by the tool"
|
|
148
|
+
},
|
|
149
|
+
{
|
|
150
|
+
"where": "tool",
|
|
151
|
+
"pattern": "EcoRI: final 568\\+56 bp vs vector 512 bp"
|
|
152
|
+
},
|
|
153
|
+
{
|
|
154
|
+
"where": "answer",
|
|
155
|
+
"pattern": "624"
|
|
156
|
+
},
|
|
157
|
+
{
|
|
158
|
+
"where": "answer",
|
|
159
|
+
"pattern": "568"
|
|
160
|
+
},
|
|
161
|
+
{
|
|
162
|
+
"where": "answer",
|
|
163
|
+
"pattern": "56\\b"
|
|
164
|
+
}
|
|
84
165
|
]
|
|
85
166
|
},
|
|
86
167
|
{
|
|
87
168
|
"id": "golden-gate",
|
|
88
169
|
"category": "cloning",
|
|
89
170
|
"tier": "full",
|
|
90
|
-
"covers": [
|
|
171
|
+
"covers": [
|
|
172
|
+
"molbio_golden_gate"
|
|
173
|
+
],
|
|
91
174
|
"instruction": "I want to assemble two fragments into my vector by Golden Gate with BsaI. The vector is a bare backbone — the region to replace is positions 451-500. Design the junctions and give me the fragments to order, then tell me the size and the overhangs.\n\nVector:\n{seq:gg_vector}\n\nFragment 1:\n{seq:gg_insert_1}\n\nFragment 2:\n{seq:gg_insert_2}",
|
|
92
175
|
"timeout_ms": 300000,
|
|
93
|
-
"expect_tools": [
|
|
176
|
+
"expect_tools": [
|
|
177
|
+
"molbio_golden_gate"
|
|
178
|
+
],
|
|
94
179
|
"assertions": [
|
|
95
|
-
{
|
|
96
|
-
|
|
97
|
-
|
|
98
|
-
|
|
180
|
+
{
|
|
181
|
+
"where": "tool",
|
|
182
|
+
"pattern": "Golden Gate assembly with BsaI \\(GGTCTC\\(1/5\\)\\): final plasmid 1226 bp \\(Δ \\+304 bp\\)"
|
|
183
|
+
},
|
|
184
|
+
{
|
|
185
|
+
"where": "tool",
|
|
186
|
+
"pattern": "2 fragment\\(s\\) to order; 3 junction\\(s\\) designed"
|
|
187
|
+
},
|
|
188
|
+
{
|
|
189
|
+
"where": "tool",
|
|
190
|
+
"pattern": "GGTCTCAAACC",
|
|
191
|
+
"note": "fragment 1's 5' flank and junction"
|
|
192
|
+
},
|
|
193
|
+
{
|
|
194
|
+
"where": "tool",
|
|
195
|
+
"pattern": "overhangs AACC / AAGC"
|
|
196
|
+
},
|
|
99
197
|
{
|
|
100
198
|
"where": "answer",
|
|
101
199
|
"pattern": "BsaI|assembl|plasmid|fragment|overhang|FASTA|deliverable",
|
|
102
200
|
"note": "the task was addressed in prose. Measured behaviour: a model that saves the final plasmid as FASTA and draws the map sometimes closes with 'both files are attached' and omits the size — the artifacts carry it. The three tool assertions above already pin the assembly arithmetic (1226 bp, 2 fragments, 3 junctions, the overhangs), so the answer check stays loose rather than punishing a delivered file."
|
|
103
201
|
}
|
|
104
202
|
]
|
|
203
|
+
},
|
|
204
|
+
{
|
|
205
|
+
"id": "silent-site",
|
|
206
|
+
"category": "cloning",
|
|
207
|
+
"tier": "full",
|
|
208
|
+
"covers": [
|
|
209
|
+
"molbio_silent_site_mutagenesis"
|
|
210
|
+
],
|
|
211
|
+
"instruction": "This 25 bp CDS contains an EcoRI site. Remove the site without changing the protein: give me the silent base change(s), the mutation texts I can pass to molbio_mutagenesis_primers, and the mutated sequence.\n\n{seq:silent_cds}",
|
|
212
|
+
"timeout_ms": 300000,
|
|
213
|
+
"expect_tools": [
|
|
214
|
+
"molbio_silent_site_mutagenesis"
|
|
215
|
+
],
|
|
216
|
+
"assertions": [
|
|
217
|
+
{
|
|
218
|
+
"where": "tool",
|
|
219
|
+
"pattern": "EcoRI \\(G\\^AATTC\\) remove: 1 site\\(s\\) before",
|
|
220
|
+
"note": "the site is found and the action is removal"
|
|
221
|
+
},
|
|
222
|
+
{
|
|
223
|
+
"where": "tool",
|
|
224
|
+
"pattern": "mutation texts: C12T",
|
|
225
|
+
"note": "the minimal silent change, TTC -> TTT (Phe)"
|
|
226
|
+
},
|
|
227
|
+
{
|
|
228
|
+
"where": "tool",
|
|
229
|
+
"pattern": "mutated sequence: ATGGCTGAATTTGGTTAAGCTTAA",
|
|
230
|
+
"note": "the protein-coding sequence with the site destroyed"
|
|
231
|
+
},
|
|
232
|
+
{
|
|
233
|
+
"where": "answer",
|
|
234
|
+
"pattern": "C12T|A9G|silent|同义",
|
|
235
|
+
"note": "the model reports a silent change"
|
|
236
|
+
}
|
|
237
|
+
]
|
|
105
238
|
}
|
|
106
239
|
]
|
|
107
240
|
}
|
|
@@ -13,78 +13,221 @@
|
|
|
13
13
|
"id": "plasmid-map-file",
|
|
14
14
|
"category": "plasmids",
|
|
15
15
|
"tier": "full",
|
|
16
|
-
"covers": [
|
|
16
|
+
"covers": [
|
|
17
|
+
"molbio_plasmid_map_file"
|
|
18
|
+
],
|
|
17
19
|
"instruction": "Draw me a circular map of pUC118.dna (in my working directory) and mark the EcoRI and HindIII cut sites. Tell me how long the plasmid is, whether it is circular, and how many annotated features it carries.",
|
|
18
20
|
"timeout_ms": 240000,
|
|
19
|
-
"expect_tools": [
|
|
20
|
-
|
|
21
|
+
"expect_tools": [
|
|
22
|
+
"molbio_plasmid_map_file"
|
|
23
|
+
],
|
|
24
|
+
"fixtures": {
|
|
25
|
+
"pUC118.dna": "test/fixtures/pUC118.dna"
|
|
26
|
+
},
|
|
21
27
|
"assertions": [
|
|
22
|
-
{
|
|
28
|
+
{
|
|
29
|
+
"where": "tool",
|
|
30
|
+
"pattern": "pUC118, 3162 bp, circular, 10 feature\\(s\\), 2 enzyme cut mark\\(s\\)"
|
|
31
|
+
},
|
|
23
32
|
{
|
|
24
33
|
"where": "answer",
|
|
25
34
|
"pattern": "3[,.]?162",
|
|
26
35
|
"note": "the plasmid length. The comma form (`3,162 bp`) is not cosmetic: a model writing a number for a human uses a thousands separator, and the first version's `/3162/` failed an otherwise perfect answer. Same rule as the Unicode minus — normalize what is formatting, grade what is content."
|
|
27
36
|
},
|
|
28
|
-
{
|
|
29
|
-
|
|
37
|
+
{
|
|
38
|
+
"where": "answer",
|
|
39
|
+
"pattern": "circular"
|
|
40
|
+
},
|
|
41
|
+
{
|
|
42
|
+
"where": "answer",
|
|
43
|
+
"pattern": "\\b10\\b|ten"
|
|
44
|
+
}
|
|
30
45
|
]
|
|
31
46
|
},
|
|
32
47
|
{
|
|
33
48
|
"id": "plasmid-map-draw",
|
|
34
49
|
"category": "plasmids",
|
|
35
50
|
"tier": "full",
|
|
36
|
-
"covers": [
|
|
51
|
+
"covers": [
|
|
52
|
+
"molbio_plasmid_map"
|
|
53
|
+
],
|
|
37
54
|
"instruction": "Render a circular plasmid map from this 512 bp sequence. Call it \"mini\", give it one feature — a CDS from position 10 to 200 labelled testCDS — and mark the EcoRI site. Then tell me what the map shows.\n\n{seq:construct}",
|
|
38
55
|
"timeout_ms": 240000,
|
|
39
|
-
"expect_tools": [
|
|
56
|
+
"expect_tools": [
|
|
57
|
+
"molbio_plasmid_map"
|
|
58
|
+
],
|
|
40
59
|
"assertions": [
|
|
41
|
-
{
|
|
42
|
-
|
|
43
|
-
|
|
44
|
-
|
|
45
|
-
{
|
|
60
|
+
{
|
|
61
|
+
"where": "tool",
|
|
62
|
+
"pattern": "mini, 512 bp, circular, 1 feature\\(s\\)"
|
|
63
|
+
},
|
|
64
|
+
{
|
|
65
|
+
"where": "tool",
|
|
66
|
+
"pattern": "1 restriction enzyme cut site\\(s\\) marked"
|
|
67
|
+
},
|
|
68
|
+
{
|
|
69
|
+
"where": "answer",
|
|
70
|
+
"pattern": "\\b512\\b"
|
|
71
|
+
},
|
|
72
|
+
{
|
|
73
|
+
"where": "answer",
|
|
74
|
+
"pattern": "testCDS|CDS"
|
|
75
|
+
},
|
|
76
|
+
{
|
|
77
|
+
"where": "answer",
|
|
78
|
+
"pattern": "EcoRI"
|
|
79
|
+
}
|
|
46
80
|
]
|
|
47
81
|
},
|
|
48
82
|
{
|
|
49
83
|
"id": "sanger-verify",
|
|
50
84
|
"category": "sequencing",
|
|
51
85
|
"tier": "full",
|
|
52
|
-
"covers": [
|
|
86
|
+
"covers": [
|
|
87
|
+
"molbio_verify_sanger"
|
|
88
|
+
],
|
|
53
89
|
"instruction": "There is a Sanger read in my working directory called read.seq. Check it against this 512 bp reference: first the reference as-is, then the same reference with the base at position 150 changed to an A. Tell me the verdict, the identity, and the difference for each case.\n\nReference:\n{seq:construct}",
|
|
54
90
|
"timeout_ms": 300000,
|
|
55
|
-
"expect_tools": [
|
|
91
|
+
"expect_tools": [
|
|
92
|
+
"molbio_verify_sanger"
|
|
93
|
+
],
|
|
56
94
|
"assertions": [
|
|
57
|
-
{
|
|
58
|
-
|
|
59
|
-
|
|
60
|
-
|
|
61
|
-
{
|
|
62
|
-
|
|
95
|
+
{
|
|
96
|
+
"where": "tool",
|
|
97
|
+
"pattern": "verdict: match[\\s\\S]*?identity 100%"
|
|
98
|
+
},
|
|
99
|
+
{
|
|
100
|
+
"where": "tool",
|
|
101
|
+
"pattern": "verdict: differences_found[\\s\\S]*?identity 99\\.67%"
|
|
102
|
+
},
|
|
103
|
+
{
|
|
104
|
+
"where": "tool",
|
|
105
|
+
"pattern": "mismatch @ref 150: A → G",
|
|
106
|
+
"note": "the planted difference, reported against the reference"
|
|
107
|
+
},
|
|
108
|
+
{
|
|
109
|
+
"where": "answer",
|
|
110
|
+
"pattern": "match"
|
|
111
|
+
},
|
|
112
|
+
{
|
|
113
|
+
"where": "answer",
|
|
114
|
+
"pattern": "99\\.6"
|
|
115
|
+
},
|
|
116
|
+
{
|
|
117
|
+
"where": "answer",
|
|
118
|
+
"pattern": "\\b150\\b"
|
|
119
|
+
}
|
|
63
120
|
]
|
|
64
121
|
},
|
|
65
122
|
{
|
|
66
123
|
"id": "crispr-guide",
|
|
67
124
|
"category": "crispr",
|
|
68
125
|
"tier": "full",
|
|
69
|
-
"covers": [
|
|
126
|
+
"covers": [
|
|
127
|
+
"molbio_grna_design"
|
|
128
|
+
],
|
|
70
129
|
"instruction": "Design the three best SpCas9 guide RNAs on this plasmid (pUC118.dna is in my working directory). For each, give me the protospacer, its strand and coordinates, the GC content and the score. Tell me how many candidate guides the scan found in total.\n\nReference sequence is the whole circular plasmid.",
|
|
71
130
|
"timeout_ms": 300000,
|
|
72
|
-
"expect_tools": [
|
|
73
|
-
|
|
131
|
+
"expect_tools": [
|
|
132
|
+
"molbio_grna_design"
|
|
133
|
+
],
|
|
134
|
+
"fixtures": {
|
|
135
|
+
"pUC118.dna": "test/fixtures/pUC118.dna"
|
|
136
|
+
},
|
|
74
137
|
"assertions": [
|
|
75
138
|
{
|
|
76
139
|
"where": "tool",
|
|
77
|
-
"pattern": "391 candidate\\(s\\) for NGG over (sequence|pUC118) \\(3162 bp\\)",
|
|
78
|
-
"note": "the scan total. The subject label is `sequence` when the model passes the sequence inline and `pUC118` when it passed a file path — both are the same scan
|
|
140
|
+
"pattern": "391 candidate\\(s\\) for SpCas9 \\[NGG PAM \\(3' of the protospacer\\)\\] over (sequence|pUC118) \\(3162 bp; target DNA, scanning both\\)",
|
|
141
|
+
"note": "the scan total and the family geometry label. The subject label is `sequence` when the model passes the sequence inline and `pUC118` when it passed a file path — both are the same scan."
|
|
142
|
+
},
|
|
143
|
+
{
|
|
144
|
+
"where": "tool",
|
|
145
|
+
"pattern": "CTGACGGGCTTGTCTGCTCC CGG \\(reverse 85-104\\)",
|
|
146
|
+
"note": "the top-ranked guide and its strand/coordinates"
|
|
147
|
+
},
|
|
148
|
+
{
|
|
149
|
+
"where": "tool",
|
|
150
|
+
"pattern": "score 100"
|
|
79
151
|
},
|
|
80
|
-
{ "where": "tool", "pattern": "CTGACGGGCTTGTCTGCTCC CGG \\(reverse 85-104\\)", "note": "the top-ranked guide and its strand/coordinates" },
|
|
81
|
-
{ "where": "tool", "pattern": "score 100" },
|
|
82
152
|
{
|
|
83
153
|
"where": "answer",
|
|
84
154
|
"pattern": "391|candidate",
|
|
85
155
|
"note": "the scan total or at least the candidate question, in prose. The guide SEQUENCE is not pinned to the answer: the tool can also write the guides to a CSV, and a model that delivers the CSV and summarises it has produced the requested artifact — the tool assertion above already proves the scan ran and what it found."
|
|
86
156
|
},
|
|
87
|
-
{
|
|
157
|
+
{
|
|
158
|
+
"where": "answer",
|
|
159
|
+
"pattern": "reverse|antisense|minus|bottom"
|
|
160
|
+
}
|
|
161
|
+
]
|
|
162
|
+
},
|
|
163
|
+
{
|
|
164
|
+
"id": "crispr-cas12a",
|
|
165
|
+
"category": "crispr",
|
|
166
|
+
"tier": "full",
|
|
167
|
+
"covers": [
|
|
168
|
+
"molbio_grna_design"
|
|
169
|
+
],
|
|
170
|
+
"instruction": "Design the best Cas12a guide on this 50 bp target. Cas12a uses a 5' TTTV PAM and a 23 nt crRNA. Give me the guide, its PAM, strand and coordinates, and the guide length — do not use SpCas9 NGG geometry.\n\nTarget:\n{seq:cas12a_target}",
|
|
171
|
+
"timeout_ms": 300000,
|
|
172
|
+
"expect_tools": [
|
|
173
|
+
"molbio_grna_design"
|
|
174
|
+
],
|
|
175
|
+
"assertions": [
|
|
176
|
+
{
|
|
177
|
+
"where": "tool",
|
|
178
|
+
"pattern": "Cas12a \\(Cpf1\\) \\[TTTV PAM \\(5' of the protospacer\\)\\]",
|
|
179
|
+
"note": "the family geometry is Cas12a's, not SpCas9's"
|
|
180
|
+
},
|
|
181
|
+
{
|
|
182
|
+
"where": "tool",
|
|
183
|
+
"pattern": "ACGTACGTACGTACGTACGTACG TTTA \\(forward 9-31\\)",
|
|
184
|
+
"note": "the 23 nt protospacer after the 5' TTTV PAM"
|
|
185
|
+
},
|
|
186
|
+
{
|
|
187
|
+
"where": "answer",
|
|
188
|
+
"pattern": "Cas12a|TTTV",
|
|
189
|
+
"note": "the model names the family/PAM it used"
|
|
190
|
+
}
|
|
191
|
+
]
|
|
192
|
+
},
|
|
193
|
+
{
|
|
194
|
+
"id": "crispr-cas13",
|
|
195
|
+
"category": "crispr",
|
|
196
|
+
"tier": "full",
|
|
197
|
+
"covers": [
|
|
198
|
+
"molbio_grna_design"
|
|
199
|
+
],
|
|
200
|
+
"instruction": "Design a Cas13a guide on this 43 bp target. Cas13a targets RNA, has a 3' H PFS preference instead of a PAM, and uses a 28 nt crRNA. Scan the sense strand; tell me the guide in RNA form, the PFS, and whether it is preferred.\n\nTarget:\n{seq:cas13_target}",
|
|
201
|
+
"timeout_ms": 300000,
|
|
202
|
+
"expect_tools": [
|
|
203
|
+
"molbio_grna_design"
|
|
204
|
+
],
|
|
205
|
+
"assertions": [
|
|
206
|
+
{
|
|
207
|
+
"where": "tool",
|
|
208
|
+
"pattern": "Cas13a \\(LwaCas13a\\) \\[no PAM \\(Cas13\\)\\]",
|
|
209
|
+
"note": "Cas13 is modelled as an RNA target with no PAM"
|
|
210
|
+
},
|
|
211
|
+
{
|
|
212
|
+
"where": "tool",
|
|
213
|
+
"pattern": "PFS preference: 3' H",
|
|
214
|
+
"note": "the preferred PFS is stated"
|
|
215
|
+
},
|
|
216
|
+
{
|
|
217
|
+
"where": "tool",
|
|
218
|
+
"pattern": "\\[RNA ACGUACGUACGUACGUACGUACGUACGU\\]",
|
|
219
|
+
"note": "the guide is reported in RNA form"
|
|
220
|
+
},
|
|
221
|
+
{
|
|
222
|
+
"where": "tool",
|
|
223
|
+
"pattern": "PFS A",
|
|
224
|
+
"note": "the observed PFS is reported"
|
|
225
|
+
},
|
|
226
|
+
{
|
|
227
|
+
"where": "answer",
|
|
228
|
+
"pattern": "Cas13|PFS|RNA",
|
|
229
|
+
"note": "the model explains the RNA/PFS distinction"
|
|
230
|
+
}
|
|
88
231
|
]
|
|
89
232
|
}
|
|
90
233
|
]
|