output results from text mining

docs/cases/PMID_25163805.txt

@@ -0,0 +1,7 @@
+[source]
+pmid = PMID:25163805
+title = Further delineation of Loeys-Dietz syndrome type 4 in a family with mild vascular involvement and a TGFB2 splicing mutation
+[diagnosis]
+disease_id = OMIM:614816
+disease_label = Loeys-Dietz syndrome 4
+[text]

docs/cases/PMID_30249733.txt

@@ -0,0 +1,26 @@
+[source]
+pmid = PMID:30249733
+title = Novel mutation in the CHST14 gene causing musculocontractural type of Ehlers-Danlos syndrome
+[diagnosis]
+disease_id = OMIM:601776
+disease_label = Ehlers-Danlos syndrome, musculocontractural type 1
+[text]
+A 3-year-old boy, born out of a third-degree consanguineous marriage was referred to us by the paediatric surgeons on
+suspicion of an underlying genetic disorder. He was being followed by them for bilateral hydronephrosis with bilateral
+pelviureteric junction obstruction (right >left). The child was born by normal vaginal delivery with a birth weight of 2.7kg.
+The length and head circumference at birth were not recorded. At birth, he was noted to have bilateral clubfeet.
+The child started sitting at around 7–8 months of age however had difficulty in standing and walking.
+At the current age of 3 years also, he is able to stand with support only for few minutes.
+In the other sectors of development like cognition and language, the child showed appropriate gain and currently is able
+to tell short stories and enjoys playing with family members. Anthropometry at the age of 3 years showed weight to be 12.6kg,
+length to be 88cm and head circumference of 47cm at 3 years of age. For the initial 1–1.5 years of life,
+the parents were mainly concerned about clubfeet in their child and were taking opinion of local practitioners for the same.
+During an episode of acute febrile illness, he was coincidently diagnosed to have hydronephrosis and in view of cryptorchidism
+noted by the examining physician was referred to our centre for evaluation and management.
+On examination, the child had facial dysmorphism in the form of synophrys, hypertelorism, down slanting palpebral fissures,
+low set ears, thin upper lip, high arched palate and prominent nasolabial folds (figures 1A and 2A–C).
+He had tapering fingers with bilaterally thin and adducted thumbs (figure 1B). The deep palmer creases were absent and
+only a few fine creases were seen. Feet showed bilateral talipes equinovarus deformity (figure 1C).
+The skin was hyperelastic and hypermobility of fingers, elbow and knee joints was noted. Generalised hypotonia was present.
+The child also had bilateral cryptorchidism. No bruises or haematomas were seen and even on repeated asking the parents denied
+any bleeding tendency.

src/main/java/org/monarchinitiative/phenopacket2prompt/Main.java

@@ -20,6 +20,7 @@ public static void main(String[] args){
                 .addSubcommand("download", new DownloadCommand())
                 .addSubcommand("prompt", new PromptCommand())
                 .addSubcommand("mine", new TextMineCommand())
+                .addSubcommand("batchmine", new TextMineCommand())
                 .addSubcommand("translate", new GptTranslateCommand())
                 ;
         cline.setToggleBooleanFlags(false);

src/main/java/org/monarchinitiative/phenopacket2prompt/cmd/BatchMineCommand.java

@@ -0,0 +1,100 @@
+package org.monarchinitiative.phenopacket2prompt.cmd;
+
+
+import org.monarchinitiative.phenol.base.PhenolRuntimeException;
+import org.monarchinitiative.phenopacket2prompt.mining.CaseBundle;
+import org.monarchinitiative.phenopacket2prompt.mining.FenominalParser;
+import org.monarchinitiative.phenopacket2prompt.model.PpktIndividual;
+import org.monarchinitiative.phenopacket2prompt.output.CorrectResult;
+import org.monarchinitiative.phenopacket2prompt.output.PpktCopy;
+import org.monarchinitiative.phenopacket2prompt.output.PromptGenerator;
+import picocli.CommandLine;
+
+import java.io.File;
+import java.util.List;
+import java.util.concurrent.Callable;
+
+@CommandLine.Command(name = "batchmine", aliases = {"B2"},
+        mixinStandardHelpOptions = true,
+        description = "Batch Text mine, Translate, and Output phenopacket and prompt")
+public class BatchMineCommand implements Callable<Integer> {
+    @CommandLine.Option(names={"-d","--data"}, description ="directory to download data (default: ${DEFAULT-VALUE})" )
+    public String datadir="data";
+
+    @CommandLine.Option(names={"-i","--inputdir"}, description ="input files (directory)" )
+    public String input = "docs/cases/"; // provide path for testing
+
+    @CommandLine.Option(names = { "-o", "--output"}, description = "Path to output file dir(default: ${DEFAULT-VALUE})")
+    private String output = "mined_out";
+
+    @CommandLine.Option(names = {"-e", "--exact"}, description = "Use exact matching algorithm")
+    private boolean useExactMatching = false;
+
+    @CommandLine.Option(names = {"--translations"},
+            description = "path to translations file")
+    private String translationsPath = "data/hp-international.obo";
+
+    @CommandLine.Option(names = {"--verbose"}, description = "show results in shell (default is to just write to file)")
+    private boolean verbose;
+
+
+
+    @Override
+    public Integer call() throws Exception {
+        File inDirectory = new File(input);
+        if (!inDirectory.isDirectory()) {
+            throw new PhenolRuntimeException("Could not find directory at " + input);
+        }
+        File hpoJsonFile = new File(datadir + File.separator + "hp.json");
+        if (! hpoJsonFile.isFile()) {
+            System.out.printf("[ERROR] Could not find hp.json file at %s\nRun download command first\n", hpoJsonFile.getAbsolutePath());
+        }
+        File translationsFile = new File(translationsPath);
+        if (! translationsFile.isFile()) {
+            System.err.printf("Could not find translations file at %s. Try download command", translationsPath);
+            return 1;
+        }
+        Utility utility = new Utility(translationsFile);
+        List<PpktIndividual> individualList =  getIndividualsFromTextMining(inDirectory,hpoJsonFile);
+        PromptGenerator spanish = utility.spanish();
+        Utility.outputPromptsInternationalMining(individualList,"es", spanish);
+        // Dutch
+        PromptGenerator dutch = utility.dutch();
+        Utility.outputPromptsInternationalMining(individualList,"nl", dutch);
+        // GERMAN
+        PromptGenerator german = utility.german();
+        Utility.outputPromptsInternationalMining(individualList,"de", german);
+        // ITALIAN
+        PromptGenerator italian = utility.italian();
+        Utility.outputPromptsInternationalMining(individualList,"it", italian);
+
+        // output file with correct diagnosis list
+        List<CorrectResult>  correctResultList =Utility.outputPromptsEnglishFromIndividuals(individualList);
+        Utility.outputCorrectTextmined(correctResultList);
+        return 0;
+    }
+
+    /**
+     * Get all of the individual objects by text mining the input files
+     * @param inDirectory Input directory. Should hold input files formatted for this project (demonstration)
+     * @param hpoJsonFile File representing hp.json
+     * @return list of individuals
+     */
+    protected List<PpktIndividual> getIndividualsFromTextMining(File inDirectory, File hpoJsonFile) {
+        FenominalParser parser = new FenominalParser(hpoJsonFile, useExactMatching);
+        List<CaseBundle> caseBundleList = Utility.getAllCaseBundlesFromDirectory(inDirectory, parser);
+        return caseBundleList.stream().map(CaseBundle::individual).toList();
+    }
+
+
+
+    private void outputTextmined(FenominalParser parser) {
+
+        List<CaseBundle> caseBundleList = Utility.getCaseBundleList(input, parser);
+        if (caseBundleList.isEmpty()) {
+            System.err.println("Could not extract cases from " + input);
+        }
+        // for now, just output one case
+        Utility.outputPromptFromCaseBundle(caseBundleList.getFirst().individual(), output);
+    }
+}