Antibody upstream sequence (AUS) repository contains all key python scripts for reproducing results in "Antibody Upstream Sequence Diversity and Its Biological Implications Revealed by Repertoire Sequencing" (doi: https://doi.org/10.1101/2020.09.02.280396). These scripts include those employed for AUS identification and filtration and those for characterizing the identified AUS from different species.
python nonhuman_aus_identification.py cloneFl alignFl primer sample leader_imgt.fasta utr_imgt.fasta isotype
The script, nonhuman_aus_identification.py, accepts seven commandline parameters and outputs a single file recording the candidate AUSs for each sample.
The comandline parameters include,
cloneFl: clone file output by MiXCRalignFl: alignment file output by MiXCRprimer: primer sequence upstream AUSsample: sample idleader_imgt.fasta: leader reference sequences from IMGT (in fasta format)utr_imgt.fasta: UTR reference sequences from IMGT (in fasta format)isotype: isotype
The output file, (i.e. sample.upstream.sequence.fasta), looks like,
Allele Group Flag Rank alleleCount novelCount Sequence leaderStart seqFreq cloneFreq readFreq seqIdentity SampleId Isotype seq_ref IMGT_leader IMGT_utr leader utr ATG_Substring
IGHV1S2*01 Homozygote upstream 1 NA 366 ACATCACACAACAATCACATCCCTCCCCTACAGAAGTCCCCAGAGCACAGCACCTCACCATGGACTGGACATGGAGGATCCTCCTCTTGGTGGCAGCAGCTACAGGCGCCCACTCC 59 0.177595628415 65 422 0.9873003487935349 MFS01 IgM ACATCACACAACAATCACATCCCTCCCCTACAGAAGTCCCCAGAGCACAGCACCTCACCATGGACTGGACATGGAGGATCCTCCTCTTGGTGGCAGCAGCTACAGGCGCCCACTCC_IGHV1S2*01 1 2 ATGGACTGGACATGGAGGATCCTCCTCTTGGTGGCAGCAGCTACAGGCGCCCACTCC ACATCACACAACAATCACATCCCTCCCCTACAGAAGTCCCCAGAGCACAGCACCTCACC GCACAGCACCTCACCATGGACTGGACATGGAGG
IGHV1S2*01 Homozygote upstream 2 NA 366 ATCACACAACAACCACATCCCTCCCCTAAAGAAGCCCCAGAGCACAGCATCTCACCATGGACTGGACCTGGAGGATCCTCCTCTTGGTGGCAGCAGCTACAGGTGCCCACTCC 56 0.106557377049 39 246 0.9752896547841291 MFS01 IgM ATCACACAACAACCACATCCCTCCCCTAAAGAAGCCCCAGAGCACAGCATCTCACCATGGACTGGACCTGGAGGATCCTCCTCTTGGTGGCAGCAGCTACAGGTGCCCACTCC_IGHV1S2*01 1 2 ATGGACTGGACCTGGAGGATCCTCCTCTTGGTGGCAGCAGCTACAGGTGCCCACTCC ATCACACAACAACCACATCCCTCCCCTAAAGAAGCCCCAGAGCACAGCATCTCACC GCACAGCATCTCACCATGGACTGGACCTGGAGG
IGHV4S9*01 Homozygote upstream 1 NA 113 ATGCTCTCTGAGAGTCACGGACATCCTGTGCAAGAACATGAAGCATCTGTGGTTCTTCCTCCTCCTGGTGGCAGCTCCCAGATGGGTCCTGTCC 37 0.106194690265 12 46 0.9230814293732988 MFS01 IgM ATGCTCTCTGAGAGTCACGGACATCCTGTGCAAGAACATGAAGCATCTGTGGTTCTTCCTCCTCCTGGTGGCAGCTCCCAGATGGGTCCTGTCC_IGHV4S9*01 1 2 ATGAAGCATCTGTGGTTCTTCCTCCTCCTGGTGGCAGCTCCCAGATGGGTCCTGTCC ATGCTCTCTGAGAGTCACGGACATCCTGTGCAAGAAC ATCCTGTGCAAGAACATGAAGCATCTGTGGTTC
.
.
.
IGHV4S10*01 Homozygote upstream 2 NA 1488 CTCTCTGAGAGTCACGGACATCCTGTGCAAGAACATGAAGCACCTGTGGTTCTTCCTCCTGCTGGTGGCAGCTCCCAGATGGGTCCTGTCC 34 0.0745967741935 111 303 0.9161023448644536 MFS01 IgM CTCTCTGAGAGTCACGGACATCCTGTGCAAGAACATGAAGCACCTGTGGTTCTTCCTCCTGCTGGTGGCAGCTCCCAGATGGGTCCTGTCC_IGHV4S10*01 1 2 ATGAAGCACCTGTGGTTCTTCCTCCTGCTGGTGGCAGCTCCCAGATGGGTCCTGTCC CTCTCTGAGAGTCACGGACATCCTGTGCAAGAAC ATCCTGTGCAAGAACATGAAGCACCTGTGGTTC
IGHV4S17*01 Homozygote upstream 1 NA 815 ACATGGGAAATGTTCTCTGAGAGTCACGGACCTCCTGGGCAAGAACATGAAGCATCTGTGGTTCTTCCTCCTCCTGGTGGCAGCTCCCAGATGGGTCCTGTCC 46 0.0895705521472 73 215 0.9133824417540037 MFS01 IgM ACATGGGAAATGTTCTCTGAGAGTCACGGACCTCCTGGGCAAGAACATGAAGCATCTGTGGTTCTTCCTCCTCCTGGTGGCAGCTCCCAGATGGGTCCTGTCC_IGHV4S17*01 1 2 ATGAAGCATCTGTGGTTCTTCCTCCTCCTGGTGGCAGCTCCCAGATGGGTCCTGTCC ACATGGGAAATGTTCTCTGAGAGTCACGGACCTCCTGGGCAAGAAC CTCCTGGGCAAGAACATGAAGCATCTGTGGTTC
IGHV4S17*01 Homozygote upstream 2 NA 815 ATGCTCTCTGAGAGTCATGGACATCCTGTGCAAGAACATGAAGCACCTGTGGTTCTTCCTCCTCCTGGTGGCAGCTCCCAGATGGGTCCTGTCC 37 0.080981595092 66 175 0.9326956172551308 MFS01 IgM ATGCTCTCTGAGAGTCATGGACATCCTGTGCAAGAACATGAAGCACCTGTGGTTCTTCCTCCTCCTGGTGGCAGCTCCCAGATGGGTCCTGTCC_IGHV4S17*01 1 2 ATGAAGCACCTGTGGTTCTTCCTCCTCCTGGTGGCAGCTCCCAGATGGGTCCTGTCC ATGCTCTCTGAGAGTCATGGACATCCTGTGCAAGAAC ATCCTGTGCAAGAACATGAAGCACCTGTGGTTC
Due to the primer difference between the in-house dataset and public dataset (PRJEB15295 and PRJNA503527), the parameters employed for identifying AUS for these datasets are different. Three variant scripts are also provided in the scripts directory. The samples for which they apply to are provided in scripts/aus_script_and_samples.csv.
python aus_score.py discovered_aus_combined.txt leader_imgt.fasta scored_aus.txt
The script, aus_score.py, implements the AUS scoring. It accepts two parameters, including the discovered AUSs combined from all samples and reference leader sequences, and outputs a single scored AUS sequence file.
python utr_sim_cal_visual.py
The script, utr_sim_cal_visual.py, implements the calculation and visualization of 5'UTR sequence similarity. It accepts the scored AUSs (i.e. human.igh.upstream.score.and.filter.q.0.50.comb.txt) and output similarity matrix files (i.e. human.igh.utr.similarity.matrix.csv) and the corresponding similarity heatmaps (i.e. human.igh.utr.similarity.matrix.jpg). The heatmap looks like,
python leader_sim_cal_visual.py
The script, leader_sim_cal_visual.py, implements the calculation and visualization of leader sequence similarity. It accepts the scored AUSs (i.e. human.igh.upstream.score.and.filter.q.0.50.comb.txt) and output similarity matrix files (i.e. human.igh.leader.similarity.matrix.csv) and the corresponding similarity heatmaps (i.e. human.igh.leader.similarity.matrix.jpg). The heatmap looks like,
python human_utr_overlap_btw_gene.py
The script, python human_utr_overlap_btw_gene.py, implement the visualization of the sharing of 5'UTR sequences between genes. It accepts the scored AUS file and the raw AUS file, and outputs heatmaps demonstrating the sharings of 5'UTR sequences between genes. The output heatmamp looks like,
python human_leader_overlap_btw_gene.py
The script, python human_leader_overlap_btw_gene.py, implement the visualization of the sharing of leader sequences between genes. It accepts the scored AUS file and the raw AUS file, and outputs heatmaps demonstrating the sharings of leader sequences between genes. The output heatmamp looks like,
python utr_sim_btw_human_and_other_species.py
The script, utr_sim_btw_human_and_other_species.py, implements the calculation and visualization of utr sequence similarity between human and other species. It accepts the scored AUSs from both human and other species (i.e. rhesus.igh.upstream.score.and.filter.q.0.50.comb.txt) and output similarity tables (i.e. human.rhesus.v.utr.identity.csv) and the corresponding scatter plots (i.e. human.other.species.v.utr.identity.corr.jpg). The heatmap looks like,
python leader_sim_btw_human_and_other_species.py
The script, leader_sim_btw_human_and_other_species.py, implements the calculation and visualization of leader sequence similarity between human and other species. It accepts the scored AUSs from both human and other species (i.e. rhesus.igh.upstream.score.and.filter.q.0.50.comb.txt) and output similarity tables (i.e. human.rhesus.v.leader.identity.csv) and the corresponding scatter plots (i.e. human.other.species.v.leader.identity.corr.jpg). The heatmap looks like,
python utr_sim_btw_human_and_other_species_boxplot.py
The script, utr_sim_btw_human_and_other_species_boxplot.py, implements the calculation and visualization of utr sequence similarity between human and other species in different gene group. It accepts the scored AUSs from both human and other species (i.e. rhesus.igh.upstream.score.and.filter.q.0.50.comb.txt), the pre-calculated pairwise gene identity matrix and the core gene list, and output similarity tables (i.e. human_other_species_utr_similarity_core_and_noncore_igh.csv) and the corresponding boxplot (i.e. utr_similarity_cmp_btw_core_noncore_for_diff_species_igh.jpg). The boxplot looks like,
python leader_sim_btw_human_and_other_species_boxplot.py
The script, leader_sim_btw_human_and_other_species_boxplot.py, implements the calculation and visualization of leader sequence similarity between human and other species in different gene groups. It accepts the scored AUSs from both human and other species (i.e. rhesus.igh.upstream.score.and.filter.q.0.50.comb.txt), the pre-calculated pairwise gene identity matrix and the core gene list, and output similarity tables (i.e. human_other_species_leader_similarity_core_and_noncore_igh.csv) and the corresponding boxplot (i.e. leader_similarity_cmp_btw_core_noncore_for_diff_species_igh.jpg). The boxplot looks like,
python v_gene_sim_btw_human_and_other_species_boxplot.py
The script, v_gene_sim_btw_human_and_other_species_boxplot.py, implements the calculation and visualization of v gene sequence similarity between human and other species in different gene groups. It accepts the pre-calculated pairwise gene identity matrix and the core gene list and output a boxplot demonstrating gene similarity distribution (i.e. v_gene_similarity_cmp_btw_core_noncore_for_diff_species_igh.jpg). The boxplot looks like,
python utr_length_cmp_btw_uorf_groups.py
The script, utr_length_cmp_btw_uorf_groups.py, implement the comparison of 5'UTR length between uORF-absent and uORF-containing groups. It accepts discovered AUSs from all enrolled species and outputs the significance of unpaired t-test for each species-chain combination and a barplot demonstrating the 5'UTR length difference between two groups. The barplot looks like,
python uorf_expr_corr.py
The script, python uorf_expr_corr.py, implement the analysis of correlation between gene expression and number of uORFs. It accepts three kinds of files as input, that includes the gene expression matrix for each sample (i.e. human.igh.gene.expr.txt), the scored AUS file, and the raw AUS file recording also the sample information. It outputs barplots demonstrating the correlation between gene expression and the number of uORFs for genes having AUSs with different number of uORFs. The output barplot looks like,
python leader_expr_corr.py
The script, python leader_expr_corr.py, implement the analysis of correlation between gene expression and leader sequences. It accepts three kinds of files as input, that includes the gene expression matrix for each sample (i.e. human.igh.gene.expr.txt), the scored AUS file, and the raw AUS file recording also the sample information. It outputs barplots demonstrating the correlation between gene expression and leader sequences for genes with leader polymorphisms. The output barplot looks like,
In-house scripts above were written in Python (v3.7). Note that a series of modules are required, which include pandas, numpy, csv, seaborn, matplotlib, warnings, multiprocessing, argparse, biopython, pyforest, subprocess, Levenshtein, scipy, sklearn, itertools, and mpl_toolkits.
For scripts performing other analyses in this work, please contact the Lead Contact, Zhenhai Zhang ([email protected]).