GEM/config.yaml at main · Elliot-Chan-120/GEM · GitHub

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
# ==================================
# Navigation
# ==================================

model_folder: 'model_folder'
database_folder: 'database'
datacore_folder: 'datacore'
pwm_folder: 'pwm_database'
dna_motifs: 'DNA_motifs'
aa_motifs: 'AA_motifs'
gene_databank: 'gene_databank'
screen_predictions: 'Screen_results'

# DATABASE FILES
clinvar_data: variant_summary.txt.gz
GRCh38_gz: GCF_000001405.40_GRCh38.p14_genomic.fna.gz
GRCh38_txt: GCF_000001405.40_GRCh38.p14_assembly_report.txt
GRCh38_fna: GCF_000001405.29_GRCh38.p14_genomic.fna

# USER GENERATED FILES FOR ML TRAINING
ref_alt_df: REF_ALT_df
context_df: CONTEXT_ref_alt_df

# Composite df - passed on to composite classes and pwm profiler for downstream analysis
composite_df: CompositeDF

# DNA dataframe
dna_profile: DNA_profile_df

# Prot dataframe
prot_profile: PROT_profile_df

# PWM dataframes
dna_pwm_profile: DNAPWM_profile_df
aa_pwm_profile: AAPWM_profile_df

full_variant_df: VARIANT_df


# ==================================
# DataFrame generation settings
# ==================================
# context / flanksize settings -> determines how much of the full sequence you want added on to each variant
flank_size: 500

# [Variant window search extensions]
DNA_buffer_region: 100
# enlarges window beyond variant where motifs and certain secondary structures / mutations will be searched
# must be smaller than flank size

AA_buffer_region: 300
# enlarges window beyond variant where protein sequences will be analyzed
# must be smaller than flank size


# ========================================
# Mutation Profile Fingerprinting Settings
# ========================================

# [DNA similarity matrix scores]
match: 1
mismatch: -1

# [K-mer size limits]
k_min: 2  # size limits on possible kmers
k_max: 7

# [Ambiguous string generation number]
variant_num: 3  # how many DNA variants we are willing to consider for ambiguous sequences

# [Functional Consequence Settings]
splice_site_window: 8
motif_window: 4

# [DNA repeat settings]
min_repeat_length: 2  # minimum size of repeat subsequences
max_repeat_length: 10
min_repeats: 5  # minimum times a subseq is repeated for its region to be considered a microsatellite

# [Protein settings]
protein_substitution_matrix: "blosum62.mat"
minimum_protein_length: 10
protein_gap_penalty: -8

# [Motif Search Settings]
motif_search_radius: 150  # reasonable length
dna_cluster_distance: 50  # base pairs
aa_cluster_distance: 30   # amino acids


# ==================================
# LookingGlass Settings
# ==================================
optimal_threshold: 0.513  # this is your trained model's optimal threshold, it is output at the end of model training

# ==================================
# Regen Settings
# ==================================
n_iterations: 50
n_copies: 1
decrease_threshold: -0.1
retain_threshold: 1
random_choices: 40
benign_goal: 0.75   # what % chance we want our resulting alleles to be Benign
scale_factor: 0.1